A Privacy-Sensitive Approach to Modeling Multi-Person Conversations
|
|
- Jeffry Atkinson
- 1 years ago
- Views:
Transcription
1 A Privacy-Sensitive Approach to Modeling Multi-Person Conversations Danny Wyatt Dept. of Computer Science University of Washington Jeff Bilmes Dept. of Electrical Engineering University of Washington Abstract In this paper we introduce a new dynamic Bayesian network that separates the speakers and their speaking turns in a multi-person conversation. We protect the speakers privacy by using only features from which intelligible speech cannot be reconstructed. The model we present combines data from multiple audio streams, segments the streams into speech and silence, separates the different speakers, and detects when other nearby individuals who are not wearing microphones are speaking. No pre-trained speaker specific models are used, so the system can be easily applied in new and different environments. We show promising results in two very different datasets that vary in background noise, microphone placement and quality, and conversational dynamics. 1 Introduction Automatically modeling people s spontaneous, face-to-face conversations is a problem of increasing interest to many different research areas. Yet there is very little data available that captures truly spontaneous speech speech recorded in situ as people go about their lives. Portable devices capable of such recording have grown in storage capacity while becoming smaller, cheaper, and more powerful. But obstacles to gathering spontaneous speech still remain, and perhaps no other obstacle is as prominent as privacy. Collecting truly spontaneous speech requires recording people in unconstrained and unpredictable situations, both public and private. There is little control over who or what may be recorded. Uninvolved parties could be recorded without their consent a scenario that, if raw audio is involved, is always unethical and often illegal. Recording spontaneous data in real-world situations will require protecting the privacy of those involved by not always storing complete audio. More specifically, any data that is saved must not allow the linguistic content of a person s speech to be reconstructed. Tanzeem Choudhury Intel Research 1100 NE 45th St., Seattle, WA Henry Kautz Dept. of Computer Science University of Rochester While that limits the analyses that can be done on the data, it does not render the data useless. A broad range of inferences can be made from privacy-sensitive features. There are many applications that would benefit from increased access to spontaneous speech data while not needing to know the content of the speech. For example, research in speech and emotion often uses only information about pitch, volume, or duration [Schuller et al., 2004]. But the data used in such research has been either acted speech [Campbell, 2000] or datasets gathered in constrained situations [Greasley et al., 1995; Douglas- Cowie et al., 2000; Ang, 2002]. Acted speech is known to poorly reflect natural emotion [Batliner et al., 2000]; and the constrained datasets are recorded in relatively unnatural settings (television shows, interviews) that are not representative of ordinary human communication. There is a demand for more natural data sets for the study of speech and emotion [Douglas-Cowie et al., 2003a]. A second example is the study of social networks. Traditional social network analysis has relied on data gathered either through surveys, which are vulnerable to known biases [Bernard et al., 1979; Marsden, 1990], or third party observers, which is costly, labor intensive, and does not scale. Recent studies have used automatically gathered data about on-line interactions [McCallum et al., 2005; Kossinet and Watts, 2006], but there are few studies involving automatically recorded face-to-face conversations despite the fact that face-to-face communication remains people s dominant mode of interaction [Baym et al., 2004]. To study social networks, it is sufficient to know only who spoke with whom, not what was said. Finally, non-linguistic aspects of spoken communication are also useful features in medical and meeting understanding applications. Speaking rate is an indicator of mental activity [Hurlburt et al., 2002] and a behavioral symptom of mania [Young et al., 1978]. Abnormal conversation dynamics are symptoms of Asperger syndrome [Wing and Gould, 1979] and autistic individuals often speak in a high-pitched voice or lack intonation [Tager-Flusberg, 1994]. In meetings, interruptions and speaking time can reveal information about 1769
2 status and dominance [Hawkins, 1991] and gender specific differences in interruptions and the consequences of those differences are active topics of research [Tannen, 1993]. None of these features require access to linguistic content, and all of these applications would benefit from increased access to natural speech data. voiced source unvoiced source gain vocal tract filter Speech 1.1 Problem Description Our specific long-term goal is to model the evolution of spontaneous face-to-face interactions within groups of individuals over extended periods of time. In order to protect the privacy of both research subjects and the non-subjects with whom they come into contact, we must ensure that the acoustic information that is saved cannot be used to reconstruct intelligible speech. The stored features must contain enough information to serve as input to models of conversational and social dynamics, but at the same time have insufficient information to reconstruct what is being said, or to positively identify individuals who are not wearing microphones. The work presented in this paper is the first step toward our goal. We present an unsupervised approach to separating speakers and their turns in a multi-person conversation, relying only on acoustic features that do not compromise privacy. The features employed are useful for modeling conversational dynamics who is speaking and how but are not sufficient for speech recognition. Our work is novel in several ways. The key contribution is a joint probabilistic model that combines streams of acoustic features from a set of individuals wearing microphones, infers when there is speech present, separates the different speakers from each other, and also detects when other individuals around them who are not equipped with microphones are speaking. It does not require pre-trained speaker specific models, and thus scales to any number of users and can be used with new speakers and in new environments. Our model can be extended to a dynamically varying number of speakers, where new audio streams come and go whenever a new person enters or exits a group. We also introduce a novel feature set that is useful for segmenting speakers and for modeling conversation attributes, but that cannot be used to transcribe the actual words spoken during a conversation. 1.2 Related Work Most of the work in modeling spoken conversations has been done in the domain of meeting understanding [NIST, 2006; Dielmann and Renals, 2004]. One of the goals in meeting understanding is speaker diarization: determining who spoke when [Reynolds and Torres-Carrasquillo, 2005]. All of the work that we have found in this domain assumes access to the full audio and that it is not necessary to remove information that can be used to transcribe speech. Much previous work has been done in linguistic conversation analysis [Ochs et al., 1996; Sacks, 1992]. Clearly, that work relies almost solely on information about the words that are spoken, and not the basic acoustics of speech. As such, the model presented in this paper is complementary to traditional conversation analysis. However, inasmuch as conversations are considered turn taking between speakers, our Figure 1: The source-filter model for speech production. model s ability to infer speaker turns can be seen as a necessary low-level enabler for higher level conversation understanding. Finally, as previously mentioned, there has been much research into recognizing emotions associated with speech [Douglas-Cowie et al., 2003b]. Many of these emotionrecognition applications may not need to know the words that are spoken. [Shriberg, 2005] mentions the importance of modeling natural speaking behavior and identifies it as a fundamental challenge for spoken language applications. 2 Speech Features and Privacy We begin by giving a very simple description of speech production based on the source-filter model [Quatieri, 2001] (see Figure 1). Most speech sounds can be modeled with two independent components: (i) the source sound generated in the glottis and (ii) the filter (the vocal tract) that shapes the spectrum of the source sound. The source can be either voiced with fundamental frequency F0 (the pitch) or unvoiced with no fundamental frequency. Prosodic information about speech (intonation, stress, and duration) is described by how the fundamental frequency and energy (volume) change during speech. The frequency response of the vocal tract particularly the resonant peaks (the formants) contains information about the actual phonemes that are the basis for words. To reproduce speech intelligibly, information on at least three formants is required [Donovan, 1996]. Any processing of the audio that removes information about the formants will ensure that intelligible speech can not be synthesized from the information that remains, and privacy will be preserved. To detect speech (and specifically, voiced speech) and model how something is being said, we extract features that contain information about the source and prosody but not about the formants. Three features that have been shown to be useful for robustly detecting voiced speech under varying noise conditions are: (i) non-initial maximum autocorrelation peak, (ii) the total number of autocorrelation peaks and (iii) relative spectral entropy [Basu, 2002] computed on 16 ms chunks of audio. Because of the periodic components of voiced speech (see Figure 1), the autocorrelation will have a small number of sharp peaks. Similarly, the relative spectral entropy between the spectrum at time t and the local mean spectrum (800 ms window) will be high for voiced speech even in the presence of indoor and outdoor noise (e.g. wind, fan). Beyond detecting the spoken regions, the system needs additional information to separate the different participants in 1770
3 G t-1 G t the complexity of the conditional probability table (CPT) we would otherwise have to model to capture dependencies between speakers. M 1 t-1 M 2 t-1 M 3 t-1 U t-1 E 1,2 t-1 E 1,3 t-1 E 2,3 t-1 V 1 t-1 V 2 t-1 V 3 t-1 O 1 t-1 O 2 t-1 O 3 t-1 A V t-1 H e t-1 M 1 t E 1,2 t E 1,3 t M 2 t E 2,3 t M 3 t V 1 t V 2 t V 3 t O 1 t O 2 t O 3 t Figure 2: DBN model for multi-person conversation the conversation. We found two features to be useful for this purpose: (i) the absolute energy, and (ii) the entropy of the energy distribution across the different microphones (described in more detail in Section 3). Summarizing, the complete list of acoustic features that must be saved for our model are: (i) non-initial maximum autocorrelation peak, (ii) the total number of autocorrelation peaks, (iii) relative spectral entropy, and (iv) energy. 3 Multi-Person Conversation Model Let us assume there are N individuals wearing microphones. Given acoustic features from these N microphones, we want to detect when one of the wearers is speaking as well as when the microphones are picking up speech from others in the area not wearing microphones. A dynamic Bayesian network (DBN) [Dean and Kanazawa, 1988] is a flexible way to combine all of the features into a unified model used to infer who is speaking when. The state factorization of DBNs makes it relatively simple to express complex dependencies between different variables in the system. Figure 2 depicts our DBN model for inferring both spoken segments as well as identifying the speaker of those segments (N =3in this example). The shaded nodes are the observed variables whose values are inputs to the system and the hidden nodes are the variables whose values are to be inferred. Each time step in the DBN corresponds to a small chunk, or frame, of audio data. In all of our experiments we used frames with a length of ms ms of overlap with the previous frame. Below we describe how the different variables are specified and the dependencies between them. Group state G t The group node G t determines who is holding the floor or taking a turn to speak. It is a discrete random variable with cardinality N +2: one state for no speaker (silent regions), one state for each of the N people wearing microphones, and one state for any other speakers not wearing microphones. The group state G t depends on G t 1, and the conditional probability P (G t G t 1 ) encodes the probability of turn transitions between speakers. At time t =0all states are equally likely: P (G 0 )=1/(N +2). The group node allows us to constrain the individual states described below, and reduces A V t U t H e t State of the individuals wearing microphones M i t The binary random variable Mt i indicates whether the i-th individual wearing a microphone is speaking. The conditional probability P (Mt i G t ) is set to be semi-deterministic: P (Mt i =1 G t = g) 1 when g = i, and 0 otherwise. This imposes the constraint that most of the time people do not talk simultaneously during a conversation. Note that it is still possible, though highly unlikely, for multiple Mt i variables to be true while G t is held to a single speaker. State of unmiked others U t Similar to Mt i, the unmiked other node U t is a binary random variable that indicates whether anyone not wearing a microphone is speaking. If there are multiple unmiked persons present, they are all modeled by this node. U t is conditioned on the group node G t and the aggregate voicing node A V t which indicates whether any microphone detected voiced speech (and is described in more detail below). The conditional probability P (U t G t,a v t =1)is defined identically to P (Mt i G t ), and P (U t =1 G t,a v t =0) 0. Voicing states Vt i and aggregate voicing A V t The voicing states Vt i are binary variables that indicate whether microphone i has recorded sound consistent with voiced human speech. The parents of Vt i are Mt i and the previous Vt 1. i Since each microphone can pick up speech from its wearer as well as other speakers nearby, the conditional probabilities of the Vt i nodes are defined as P (Vt i = 1 Mt i =1) 1 and P (Vt i =1 Mt i =0)=0.5. In other words, if person i is speaking it is highly likely her microphone will record voiced speech, and if she is not speaking there is a uniform probability that her microphone will record voiced speech. The node A V t is an aggregate voicing node that is the deterministic logical OR of all the Vt i nodes. We describe below how A V t helps distinguish other individuals speaking from silent regions. Observations O i t, E i,j t, and H e t The observed variables obtained from the acoustic features of the N microphones are included at various points in the DBN as children of Mt i, U t, and Vt i. Ot i is a three-dimensional variable that includes the three features previously mentioned as having been useful for detecting voiced speech (non-initial maximum autocorrelation peak, number of autocorrelation peaks, and relative spectral entropy). P (Ot i = o Vt i = v) is modeled by a 3D Gaussian with full covariance matrix. The P (Ot V i t i ) parameters are learned from a set of labeled data (where Vt i is given) containing speakers who are not present in any of the data we evaluated here. Learning these features in this manner has been shown to be speaker-independent and robust across different environmental conditions [Choudhury and Basu, 2004]. E i,j t is a two-dimensional variable containing the log energies of microphones i and j averaged over a 333 ms window centered at time t. The conditional distribution 1771
4 log energy B log entropy Speaker B Speaker A Neither Both log energy A Figure 3: Pairwise log energies. Speaker A Unmiked Speaker Speaker B Speaker A Time (seconds) Figure 4: Log entropy of energy distribution across microphones. P (E i,j t Mt i,m j t ) is modeled with a full covariance 2D Gaussian. This pairwise energy feature associates voiced regions with their speaker. If person i speaks at time t, then her energy should be higher than j s (and vice versa). When both i and j speak, both of their microphones have high energy and when neither speaks both of their microphones have low energy. Figure 3 illustrates an example of this. Ht e is the entropy of the log-energy distribution across all N microphones. This feature is useful for determining whether voiced regions come from a speaker not wearing a microphone. When a person wearing a microphone speaks, his microphone will be significantly louder than the others and the entropy will be low. When a person not wearing a microphone speaks, her energy will be spread more uniformly across all microphones and the entropy will be high. Figure 4 illustrates an example of this. Ht e is computed as follows. First, all microphones energies are normalized to a distribution: P e (τ) = j ei(τ) ej(τ). Then H τ is computed as the entropy of P e (τ). Finally, Ht e is the average of H τ over a 333 ms window centered at time t. We use the log of Ht e and model that with a 1D Gaussian random variable conditioned on U t. Although Ht e is a useful feature for distinguishing whether speech comes from a person wearing a microphone or from one who is not, it does not help distinguish between when an unmiked person is speaking and when no one is speaking. Entropy is high in both of those cases. However, information about the voicing states Vt i taken together with the entropy can distinguish those situations. If there is voicing and the entropy is high then it is likely that someone else (i.e. someone not wearing a microphone) is speaking. It is here that the aggregate voicing node A V t is useful. We define P (U t G t = u, A V t =1) 1 (where u is the state of node G t that indicates an unmiked person is speaking), and P (U t G t,a V t ) 0 if G t u or if A V t =0. Loosely, this means that the model will only infer an unmiked speaker if at least one microphone picked up voiced speech and that speech cannot be assigned to any of the miked speakers. 3.1 Parameter Learning and Inference Learning is done in an entirely unsupervised manner using expectation maximization (EM). Unsupervised learning is important for this application, given the privacy constraints associated with recording spontaneous speech. Raw audio will not be available for labeling speaker-specific data, so the model must be able to fit itself to unlabeled data. However, with a large number of parameters, EM can often converge to values that do not result in accurate inferences. To prevent this, we clamp most of the above parameters to their pre-defined or pre-trained values. Indeed, only the Gaussians associated with the energy-based observations (P (E i,j t M i,m j ) and P (Ht e U t )) are learned during EM. (As mentioned, the Gaussians associated with the voicing observations are pre-trained in a speaker-independent way.) All of the transition probabilities and semi-deterministic conditional probabilities are fixed at predefined values. We did experiment with leaving more parameters free and the resulting inferences were much less accurate. Once the free parameters are learned with EM, exact inference is done using the junction tree algorithm. During decoding, we infer the most likely state sequence for the group node G, speaker nodes Mt i and U t, and voicing nodes Vt i. We use the Graphical Models Toolkit (GMTK) for all of our learning and inference [Bilmes and Zweig, 2002]. 4 Experiments and Results Experimental evaluations were performed on two datasets: (i) the publicly available scripted meeting corpus (M4) from IDIAP [McCowan et al., 2003] and (ii) a labeled dataset of natural interactions that we collected. The M4 corpus contains 27 four person meetings, each of them about five minutes long. The dataset has audio recordings from 12 microphones one 8 microphone array and 4 lapel microphones. The speakers in each recording followed a script for certain meeting-wide activities (e.g., discussion, argument, monologue) but were not told what to say. In our experiments, we used data only from the lapel microphones as it most closely resembles our data collection setup. Overall, this dataset is quite clean and does not have much background noise. We evaluated the performance of our model on 13 randomly selected meetings from this corpus. The dataset we collected is much more challenging. It has a significant amount of background noise and distant speech. There are 6 conversations collected in 4 different locations: a meeting room, an elevator, a hallway, and a loud and noisy atrium. The speakers were told where to go but not what to talk about. They are all friends and had no trouble filling 1772
5 mics frame err. DER prec. recall Table 1: Results for the M4 corpus. All meetings had 4 participants. Figure 5: Two people wearing our portable recording equipment. The sensing unit is at their right shoulders. the time with spontaneous conversation. For recording, we used an inexpensive condenser microphone, which is part of a multi-modal sensing unit (dimensions: 60 mm x 30 mm x 25 mm) [Welbourne et al., 2005]. The sensing unit is clipped to the strap of a small over the shoulder bag. The unit sits near the upper right shoulder, but can move so the microphone is not always at a fixed location from the mouth. A PDA in the bag records the audio data. Figure 5 shows two people wearing the equipment. Unlike the M4 data, where all lapel microphones are tethered to the same recording computer, each person in our data carries with her all the equipment needed for recording. Thus, our participants can move about independently and interact in a more natural manner. To evaluate the speaker segmentation performance, we learned the unclamped parameters of the model in an unsupervised manner for each meeting independently. Once the learning was done, we inferred the most likely state sequence for the group state node G t and speaker nodes Mt i and U t. We compute four evaluation metrics that compare the inferred value of G t to ground truth. (i) The per frame error rate is the fraction of frames in which the value of G t does not match the ground truth speaker. We do not consider frames that have more than one ground truth speaker. (ii) The diarization error rate (DER) is a standard metric used by NIST [NIST, 2006] to measure the performance of speaker segmentation systems. It is a relaxed version of frame error rate that merges pauses shorter than 0.3 s long and ignores 0.25 s of data around a change in speaker. These relaxations account for perceptual difficulties in labeling speech at such a fine time granularity. (iii) Precision is the fraction of the total number of inferred speaker frames that are correct. (iv) Recall is the fraction of truly spoken frames for which any speaker is inferred (in other words, the accuracy of basic speech-detection). The results for the M4 corpus are in Table 1 and the results for our dataset are in Tables 2(a) and 2(b). In both datasets, all participants wear microphones. To test the performance of our model with unmiked speakers, we selectively ignored the data from some participants microphones. Results shown for speakers mics frame err. DER prec. recall (a) Quiet environments. speakers mics frame err. DER prec. recall (b) Noisy environments Table 2: Results on our data. cases with fewer microphones than speakers are the averages of results for all permutations of that number of microphones across that number of speakers. The first thing to note is that the DER scores on the M4 data are comparable with current speaker diarization results is currently the best DER (achieved with features that do not preserve privacy) for meeting data [NIST, 2006]. (Unfortunately the dataset used in that evaluation is not generally available, so we cannot compare our results directly.) To the best of our knowledge, there is only one other published diarization result for the M4 corpus [Ajmera et al., 2004]. That technique has better frame error rates (7.4%) but lower precision and recall (their reported average of the two is 80.8, our average of the two is 88.5) thus making for an inconclusive comparison. That technique also used features (low order cepstral coefficients) that contain information about the words spoken, so it does not protect privacy. Our error rates are significantly better on the M4 dataset than on the dataset we collected. Our dataset, however, has substantially more difficult characteristics than M4. Our data includes significantly more background noise, and the conversations are more fluent and fast paced with much more speaker overlap. For example, the M4 data s mean turn duration is 6.5 s (median: 2.5 s). For our dataset the mean turn duration is 1.52 s (median: 1.1 s). Clearly, more work remains to be done in handling noisy environments and conversations with faster (or even variable) pacing, but the results are still quite promising. 1773
6 Error Rate A B C D E Time Figure 6: Model history 5 Discussion Our goal of separating the different speakers in conversations and identifying when they speak while not preserving information from which intelligible speech can be reconstructed led us to explore many different variations of the model proposed in this paper. Having the correct graphical structure and acoustic features is critical in solving this problem. A number of variations were attempted along the way. This is depicted in the development history plot shown in Figure 6. We started by using only pairwise relative energy ratios along with the voicing features for our observations (A). This performed quite poorly, failing in cases where no one was speaking and small absolute changes in energy caused large changes in their ratios. In (B), we modeled the full distribution of pairwise energies which significantly improved our error rates. Next, we experimented with the effect of making the individual speaker variables Mt i deterministically dependent on G t which worsened things (C) and lead to the semideterministic CPT we ultimately used. In (D) and (E) we tried adding various energy statistics as children of U t to determine when unmiked persons were speaking. For (D) we added the maximum microphone energy, and in (E) we added the mean and variance of all the energies, but these models were not appreciably better. Removing these energy observations and using an aggregate voicing indicator A V t alone in model (F), did help (F) is 3% lower than (B). Model (G) was an attempt to add back the mean and variance of energy, but those features (even after normalization) continued to hurt performance. This suggests that those features are not sufficiently discriminative which causes EM to converge to unhelpful parameter values. Lastly, model (H) uses the smoothed log entropy of the energy distribution, as described earlier in Section 3. F G H 6 Future Work There are many ways that our model could be extended. Currently, it requires at least two microphones, and we are exploring ways to allow it to work with only one microphone. And while the model is theoretically capable of inferring overlapping speakers, the semi-deterministic conditional probabilities from the G t node prevent it from reliably doing so. We are exploring ways of adding a switching node [Bilmes, 2000] to explicitly represent changes in speaker turns as well as interruptions and interjections. Additionally, we have not experimented with very many different values of the pre-defined parameters. It may be possible to learn some of these from other data, as we do with the voicing parameters, or to allow EM to learn them within given bounds. Learning better turn transition probabilities will probably help our results the most, since turn lengths seem to vary with conversation type. We also plan to apply this model to the analysis of a much larger data set. We have recently collected over 4,400 hours of data from 24 subjects over the course of 9 months. The subjects wore our data collection equipment and all feature extraction was done in real-time on the PDA. (The data also contains features beyond those listed in this paper, but none from which intelligible speech can be reconstructed.) We hope to use our model (along with other techniques) to study the evolution of the subjects conversational styles and social network. To the best of our knowledge, this is the most detailed study of face-to-face social interactions ever done using automatically gathered data. We would not have been able to collect this data without the privacy guarantees provided by our feature set. 7 Conclusion We have presented a DBN and privacy-sensitive feature set that are capable of inferring who was speaking when in a conversation. The feature set does not include any information that could be used to reconstruct intelligible speech. In clean data, its performance is comparable to that of systems that use much richer features from which the original speech can be reconstructed. Even though other features useful for speech synthesis and recognition could have been used, we believe there is a huge advantage to protecting privacy. In the long run, this will allow us to collect and model more interesting and spontaneous conversations and extend this work to capture richer conversation dynamics and handle varying numbers of speakers and overlaps. Acknowledgements This work was supported by National Science Foundation grant IIS References [Ajmera et al., 2004] J. Ajmera, G. Lathoud, and L. Mc- Cowan. Clustering and segmenting speakers and their locations in meetings. In Proceedings of ICASSP, [Ang, 2002] J. Ang. Prosody-based automatic detection of annoyance and frustration in human-computer dialog. In Proceedings of ICSLP, [Basu, 2002] S. Basu. Conversational Scene Analysis. PhD Thesis, MIT, [Batliner et al., 2000] A. Batliner, K. Fisher, R. Huber, J. Spilker, and E. Noth. Desperately seeking emotions or: Actors, wizards and human beings. In Proceedings of the ISCA ITRW on Speech and Emotion, [Baym et al., 2004] N. Baym, Y. B. Zhang, and M. C. Lin. Social interactions across media: Interpersonal communication on the internet, telephone and face-to-face. New Media and Society, 6: ,
7 [Bernard et al., 1979] H. Bernard, P. Killworth, and L. Sailer. Informant accuracy in social network data iv. Social Networks, 2: , [Bilmes and Zweig, 2002] J. Bilmes and G. Zweig. The graphical models toolkit: An open source software system for speech and time-series processing. In Proceedings of ICASSP, [Bilmes, 2000] J. A. Bilmes. Dynamic bayesian multinets. In Proceedings of UAI, [Campbell, 2000] N. Campbell. Databases of emotional speech. In Proceedings of the ISCA ITRW on Speech and Emotion, [Choudhury and Basu, 2004] T. Choudhury and S. Basu. Modeling conversation dynamics as a mixed memory markov process. In Proceedings of NIPS, [Dean and Kanazawa, 1988] T. Dean and K. Kanazawa. Probabilistic temporal reasoning. In Proceedings of AAAI, [Dielmann and Renals, 2004] A. Dielmann and S. Renals. Multi-stream segmentation of meetings. In Proceedings of the IEEE Workshop on Multimedia Signal Processing, [Donovan, 1996] R. Donovan. Trainable Speech Synthesis. PhD thesis, Cambridge University, [Douglas-Cowie et al., 2000] E. Douglas-Cowie, R. Cowie, and M. Schroeder. A new emotion database: Considerations, sources and scope. In Proceedings of the ISCA ITRW on Speech and Emotion, [Douglas-Cowie et al., 2003a] E. Douglas-Cowie, N. Campbell, R. Cowie, and P. Roach. Emotional speech: Towards a new generation of databases. Speech Communication, 40(1 2):33 60, [Douglas-Cowie et al., 2003b] E. Douglas-Cowie, R. Cowie, and N. Campbell. Speech and emotion. Speech Communication, 40(1 2):1 3, [Greasley et al., 1995] P. Greasley, J. Setter, M. Waterman, C. Sherrard, P. Roach, S. Arnfield, and D. Horton. Representation of prosodic and emotional features in a spoken language database. In Proceedings of the XIIIth ICPhS, [Hawkins, 1991] K. Hawkins. Some consequences of deep interruption in task-oriented communication. Journal of Language and Social Psychology, 10: , [Hurlburt et al., 2002] R. T. Hurlburt, M. Koch, and C. L. Heavey. Descriptive experience sampling demonstrates the connection of thinking to externally observable behavior. Cognitive Therapy and Research, 26(1): , [Kossinet and Watts, 2006] G. Kossinet and D. J. Watts. Empirical analysis of an evolving social network. Science, 311:88 90, [Marsden, 1990] P. V. Marsden. Network data and measurement. Annual Review of Sociology, 16: , [McCallum et al., 2005] A. McCallum, A. Corrada- Emmanuel, and X. Wang. Topic and role discovery in social networks. In Proceedings of IJCAI, [McCowan et al., 2003] I. McCowan, S. Bengio, D. Gatica- Perez, F. Lathoud, D. Moore, P. Wellner, and H. Bourlard. Modeling human interactions in meetings. In Proceedings of ICASSP, [NIST, 2006] NIST. NIST rich transcription evaluations [Ochs et al., 1996] E. Ochs, E. A. Schegloff, and S. A. Thompson, editors. Interaction and Grammar. Cambridge, [Quatieri, 2001] T. Quatieri. Discrete-Time Speech Signal Processing: Principles and Practice. Prentice Hall, 1st edition edition, [Reynolds and Torres-Carrasquillo, 2005] D. A. Reynolds and P. Torres-Carrasquillo. Approaches and applications of audio diarization. In Proceedings of ICASSP, [Sacks, 1992] H. Sacks. Lectures on Conversation. Blackwell, [Schuller et al., 2004] B. Schuller, G. Rigoll, and M. Lang. Speech emotion recognition combining acoustic features and linguistic information in a hybrid support vector machine belief network architecture. In Proceedings of ICASSP, [Shriberg, 2005] E. Shriberg. Spontaneous speech: How people really talk and why engineers should care. In Proceedings of Eurospeech, [Tager-Flusberg, 1994] H. Tager-Flusberg. Dissociations in form and function in the acquisition of language by autistic children. In H. Tager-Flusberg, editor, Constraints on language acquisition: Studies of atypical children, pages Lawrence Earlbaum, [Tannen, 1993] D. Tannen. Interpreting interruption in conversation. In Gender and Discourse, pages Oxford, [Welbourne et al., 2005] E. Welbourne, J. Lester, A. LaMarca, and G. Borriello. Mobile context inference using low-cost sensors. In Proceedings of LoCA, [Wing and Gould, 1979] L. Wing and J. Gould. Severe impairment of social interaction and associated abnormalities in children: Epidemiology and classification. Journal of Autism and Developmental Disorders, 9:11 29, [Young et al., 1978] R. C. Young, J. T. Biggs, V. E. Ziegler, and D. A. Meyer. A rating scale for mania: Reliability, validity and sensitivity. British Journal of Psychiatry, 133: ,
A Hybrid System for Audio Segmentation and Speech endpoint Detection of Broadcast News
A Hybrid System for Audio Segmentation and Speech endpoint Detection of Broadcast News Maria Markaki 1, Alexey Karpov 2, Elias Apostolopoulos 1, Maria Astrinaki 1, Yannis Stylianou 1, Andrey Ronzhin 2
ONLINE SPEAKER DIARIZATION USING ADAPTED I-VECTOR TRANSFORMS. Weizhong Zhu and Jason Pelecanos. IBM Research, Yorktown Heights, NY 10598, USA
ONLINE SPEAKER DIARIZATION USING ADAPTED I-VECTOR TRANSFORMS Weizhong Zhu and Jason Pelecanos IBM Research, Yorktown Heights, NY 1598, USA {zhuwe,jwpeleca}@us.ibm.com ABSTRACT Many speaker diarization
Learning Situation Models for Providing Context-Aware Services
Learning Situation Models for Providing Context-Aware Services O. Brdiczka, J.L. Crowley, and P. Reignier Project PRIMA, INRIA Rhône-Alpes, Montbonnot, France {brdiczka, crowley, reignier}@inrialpes.fr
FILTER BANK FEATURE EXTRACTION FOR GAUSSIAN MIXTURE MODEL SPEAKER RECOGNITION
FILTER BANK FEATURE EXTRACTION FOR GAUSSIAN MIXTURE MODEL SPEAKER RECOGNITION James H. Nealand, Alan B. Bradley, & Margaret Lech School of Electrical and Computer Systems Engineering, RMIT University,
Table 1: Classification accuracy percent using SVMs and HMMs
Feature Sets for the Automatic Detection of Prosodic Prominence Tim Mahrt, Jui-Ting Huang, Yoonsook Mo, Jennifer Cole, Mark Hasegawa-Johnson, and Margaret Fleck This work presents a series of experiments
Tencent AI Lab Rhino-Bird Visiting Scholar Program. Research Topics
Tencent AI Lab Rhino-Bird Visiting Scholar Program Research Topics 1. Computer Vision Center Interested in multimedia (both image and video) AI, including: 1.1 Generation: theory and applications (e.g.,
Detecting Group Turns of Speaker Groups in Meeting Room Conversations Using Audio-Video Change Scale-Space
University of South Florida Scholar Commons Graduate Theses and Dissertations Graduate School 6-30-2010 Detecting Group Turns of Speaker Groups in Meeting Room Conversations Using Audio-Video Change Scale-Space
Foreign Accent Classification
Foreign Accent Classification CS 229, Fall 2011 Paul Chen pochuan@stanford.edu Julia Lee juleea@stanford.edu Julia Neidert jneid@stanford.edu ABSTRACT We worked to create an effective classifier for foreign
Abstract. 1 Introduction. 2 Background
Automatic Spoken Affect Analysis and Classification Deb Roy and Alex Pentland MIT Media Laboratory Perceptual Computing Group 20 Ames St. Cambridge, MA 02129 USA dkroy, sandy@media.mit.edu Abstract This
Gender Classification Based on FeedForward Backpropagation Neural Network
Gender Classification Based on FeedForward Backpropagation Neural Network S. Mostafa Rahimi Azghadi 1, M. Reza Bonyadi 1 and Hamed Shahhosseini 2 1 Department of Electrical and Computer Engineering, Shahid
Speaker Recognition Using MFCC and GMM with EM
RESEARCH ARTICLE OPEN ACCESS Speaker Recognition Using MFCC and GMM with EM Apurva Adikane, Minal Moon, Pooja Dehankar, Shraddha Borkar, Sandip Desai Department of Electronics and Telecommunications, Yeshwantrao
Zaki B. Nossair and Stephen A. Zahorian Department of Electrical and Computer Engineering Old Dominion University Norfolk, VA, 23529
SMOOTHED TIME/FREQUENCY FEATURES FOR VOWEL CLASSIFICATION Zaki B. Nossair and Stephen A. Zahorian Department of Electrical and Computer Engineering Old Dominion University Norfolk, VA, 23529 ABSTRACT A
L18: Speech synthesis (back end)
L18: Speech synthesis (back end) Articulatory synthesis Formant synthesis Concatenative synthesis (fixed inventory) Unit-selection synthesis HMM-based synthesis [This lecture is based on Schroeter, 2008,
Accent Classification
Accent Classification Phumchanit Watanaprakornkul, Chantat Eksombatchai, and Peter Chien Introduction Accents are patterns of speech that speakers of a language exhibit; they are normally held in common
THIRD-ORDER MOMENTS OF FILTERED SPEECH SIGNALS FOR ROBUST SPEECH RECOGNITION
THIRD-ORDER MOMENTS OF FILTERED SPEECH SIGNALS FOR ROBUST SPEECH RECOGNITION Kevin M. Indrebo, Richard J. Povinelli, and Michael T. Johnson Dept. of Electrical and Computer Engineering, Marquette University
L12: Template matching
Introduction to ASR Pattern matching Dynamic time warping Refinements to DTW L12: Template matching This lecture is based on [Holmes, 2001, ch. 8] Introduction to Speech Processing Ricardo Gutierrez-Osuna
Pavel Král and Václav Matoušek University of West Bohemia in Plzeň (Pilsen), Czech Republic pkral
EVALUATION OF AUTOMATIC SPEAKER RECOGNITION APPROACHES Pavel Král and Václav Matoušek University of West Bohemia in Plzeň (Pilsen), Czech Republic pkral matousek@kiv.zcu.cz Abstract: This paper deals with
Automatic Phonetic Alignment and Its Confidence Measures
Automatic Phonetic Alignment and Its Confidence Measures Sérgio Paulo and Luís C. Oliveira L 2 F Spoken Language Systems Lab. INESC-ID/IST, Rua Alves Redol 9, 1000-029 Lisbon, Portugal {spaulo,lco}@l2f.inesc-id.pt
Sequence Discriminative Training;Robust Speech Recognition1
Sequence Discriminative Training; Robust Speech Recognition Steve Renals Automatic Speech Recognition 16 March 2017 Sequence Discriminative Training;Robust Speech Recognition1 Recall: Maximum likelihood
An Improvement of robustness to speech loudness change for an ASR system based on LC-RC features
An Improvement of robustness to speech loudness change for an ASR system based on LC-RC features Pavel Yurkov, Maxim Korenevsky, Kirill Levin Speech Technology Center, St. Petersburg, Russia Abstract This
Discriminative Learning of Feature Functions of Generative Type in Speech Translation
Discriminative Learning of Feature Functions of Generative Type in Speech Translation Xiaodong He Microsoft Research, One Microsoft Way, Redmond, WA 98052 USA Li Deng Microsoft Research, One Microsoft
COMPARISON OF EVALUATION METRICS FOR SENTENCE BOUNDARY DETECTION
COMPARISON OF EVALUATION METRICS FOR SENTENCE BOUNDARY DETECTION Yang Liu Elizabeth Shriberg 2,3 University of Texas at Dallas, Dept. of Computer Science, Richardson, TX, U.S.A 2 SRI International, Menlo
Classification with Deep Belief Networks. HussamHebbo Jae Won Kim
Classification with Deep Belief Networks HussamHebbo Jae Won Kim Table of Contents Introduction... 3 Neural Networks... 3 Perceptron... 3 Backpropagation... 4 Deep Belief Networks (RBM, Sigmoid Belief
Discriminative Learning of Feature Functions of Generative Type in Speech Translation
Discriminative Learning of Feature Functions of Generative Type in Speech Translation Xiaodong He Microsoft Research, One Microsoft Way, Redmond, WA 98052 USA Li Deng Microsoft Research, One Microsoft
Speech Synthesizer for the Pashto Continuous Speech based on Formant
Speech Synthesizer for the Pashto Continuous Speech based on Formant Technique Sahibzada Abdur Rehman Abid 1, Nasir Ahmad 1, Muhammad Akbar Ali Khan 1, Jebran Khan 1, 1 Department of Computer Systems Engineering,
STA 414/2104 Statistical Methods for Machine Learning and Data Mining
STA 414/2104 Statistical Methods for Machine Learning and Data Mining Radford M. Neal, University of Toronto, 2014 Week 1 What are Machine Learning and Data Mining? Typical Machine Learning and Data Mining
Modulation frequency features for phoneme recognition in noisy speech
Modulation frequency features for phoneme recognition in noisy speech Sriram Ganapathy, Samuel Thomas, and Hynek Hermansky Idiap Research Institute, Rue Marconi 19, 1920 Martigny, Switzerland Ecole Polytechnique
Speech Emotion Recognition Using Support Vector Machine
Speech Emotion Recognition Using Support Vector Machine Yixiong Pan, Peipei Shen and Liping Shen Department of Computer Technology Shanghai JiaoTong University, Shanghai, China panyixiong@sjtu.edu.cn,
Phoneme Recognition Using Deep Neural Networks
CS229 Final Project Report, Stanford University Phoneme Recognition Using Deep Neural Networks John Labiak December 16, 2011 1 Introduction Deep architectures, such as multilayer neural networks, can be
Performance Analysis of Spoken Arabic Digits Recognition Techniques
JOURNAL OF ELECTRONIC SCIENCE AND TECHNOLOGY, VOL., NO., JUNE 5 Performance Analysis of Spoken Arabic Digits Recognition Techniques Ali Ganoun and Ibrahim Almerhag Abstract A performance evaluation of
Speaker Recognition Using Vocal Tract Features
International Journal of Engineering Inventions e-issn: 2278-7461, p-issn: 2319-6491 Volume 3, Issue 1 (August 2013) PP: 26-30 Speaker Recognition Using Vocal Tract Features Prasanth P. S. Sree Chitra
HMM-Based Emotional Speech Synthesis Using Average Emotion Model
HMM-Based Emotional Speech Synthesis Using Average Emotion Model Long Qin, Zhen-Hua Ling, Yi-Jian Wu, Bu-Fan Zhang, and Ren-Hua Wang iflytek Speech Lab, University of Science and Technology of China, Hefei
Automatic Text Summarization for Annotating Images
Automatic Text Summarization for Annotating Images Gediminas Bertasius November 24, 2013 1 Introduction With an explosion of image data on the web, automatic image annotation has become an important area
Object Identication in Dynamic Images Based on the Memory-Prediction Theory of Brain Function
Object Identication in Dynamic Images Based on the Memory-Prediction Theory of Brain Function Marek Bundzel, PhD. Technical University Košice Dept. of Cybernetics and Artificial Intelligence marek.bundzel@tuke.sk
ANNOTATION AND DETECTION OF CONFLICT ESCALATION IN POLITICAL DEBATES
ANNOTATION AND DETECTION OF CONFLICT ESCALATION IN POLITICAL DEBATES Samuel Kim 1,2, Fabio Valente 2 and Alessandro Vinciarelli 2, 1 DSP Lab., Yonsei University, Seoul, Korea 2 Idiap Research Institute,
Learning words from sights and sounds: a computational model. Deb K. Roy, and Alex P. Pentland Presented by Xiaoxu Wang.
Learning words from sights and sounds: a computational model Deb K. Roy, and Alex P. Pentland Presented by Xiaoxu Wang Introduction Infants understand their surroundings by using a combination of evolved
Low-Delay Singing Voice Alignment to Text
Low-Delay Singing Voice Alignment to Text Alex Loscos, Pedro Cano, Jordi Bonada Audiovisual Institute, Pompeu Fabra University Rambla 31, 08002 Barcelona, Spain {aloscos, pcano, jboni }@iua.upf.es http://www.iua.upf.es
A Flexible Framework for Key Audio Effects Detection and Auditory Context Inference
1026 IEEE TRANSACTIONS ON AUDIO, SPEECH, AND LANGUAGE PROCESSING, VOL. 14, NO. 3, MAY 2006 A Flexible Framework for Key Audio Effects Detection and Auditory Context Inference Rui Cai, Lie Lu, Member, IEEE,
PERFORMANCE OF SRI'S DECIPHER TM SPEECH RECOGNITION SYSTEM ON DARPA'S CSR TASK 1. ABSTRACT 4. PORTING DECIPHER TM TO THE CSR TASK 2.
PERFORMANCE OF SRI'S DECIPHER TM SPEECH RECOGNITION SYSTEM ON DARPA'S CSR TASK Hy Murveit, John Butzberger, and Mitch Weintraub SRI International Speech Research and Technology Program Menlo Park, CA,
L16: Speaker recognition
L16: Speaker recognition Introduction Measurement of speaker characteristics Construction of speaker models Decision and performance Applications [This lecture is based on Rosenberg et al., 2008, in Benesty
Using Word Confusion Networks for Slot Filling in Spoken Language Understanding
INTERSPEECH 2015 Using Word Confusion Networks for Slot Filling in Spoken Language Understanding Xiaohao Yang, Jia Liu Tsinghua National Laboratory for Information Science and Technology Department of
ROBUST SPEECH RECOGNITION FROM RATIO MASKS. {wangzhon,
ROBUST SPEECH RECOGNITION FROM RATIO MASKS Zhong-Qiu Wang 1 and DeLiang Wang 1, 2 1 Department of Computer Science and Engineering, The Ohio State University, USA 2 Center for Cognitive and Brain Sciences,
Robust DNN-based VAD augmented with phone entropy based rejection of background speech
INTERSPEECH 2016 September 8 12, 2016, San Francisco, USA Robust DNN-based VAD augmented with phone entropy based rejection of background speech Yuya Fujita 1, Ken-ichi Iso 1 1 Yahoo Japan Corporation
Sentiment Analysis of Speech
Sentiment Analysis of Speech Aishwarya Murarka 1, Kajal Shivarkar 2, Sneha 3, Vani Gupta 4,Prof.Lata Sankpal 5 Student, Department of Computer Engineering, Sinhgad Academy of Engineering, Pune, India 1-4
learn from the accelerometer data? A close look into privacy Member: Devu Manikantan Shila
What can we learn from the accelerometer data? A close look into privacy Team Member: Devu Manikantan Shila Abstract: A handful of research efforts nowadays focus on gathering and analyzing the data from
Automatic identification of individual killer whales
Automatic identification of individual killer whales Judith C. Brown a) Department of Physics, Wellesley College, Wellesley, Massachusetts 02481 and Media Laboratory, Massachusetts Institute of Technology,
Session 1: Gesture Recognition & Machine Learning Fundamentals
IAP Gesture Recognition Workshop Session 1: Gesture Recognition & Machine Learning Fundamentals Nicholas Gillian Responsive Environments, MIT Media Lab Tuesday 8th January, 2013 My Research My Research
Clustered Knowledge Tracing
Clustered nowledge Tracing Zachary A. Pardos, Shubhendu Trivedi, Neil T. Heffernan, Gábor N. Sárközy Department of Computer Science, Worcester Polytechnic Institute, United States {zpardos,s_trivedi,nth,gsarkozy}@cs.wpi.edu
A comparison between human perception and a speaker verification system score of a voice imitation
PAGE 393 A comparison between human perception and a speaker verification system score of a voice imitation Elisabeth Zetterholm, Mats Blomberg 2, Daniel Elenius 2 Department of Philosophy & Linguistics,
Learning Methods in Multilingual Speech Recognition
Learning Methods in Multilingual Speech Recognition Hui Lin Department of Electrical Engineering University of Washington Seattle, WA 98125 linhui@u.washington.edu Li Deng, Jasha Droppo, Dong Yu, and Alex
Dynamic Vocal Tract Length Normalization in Speech Recognition
Dynamic Vocal Tract Length Normalization in Speech Recognition Daniel Elenius, Mats Blomberg Department of Speech Music and Hearing, CSC, KTH, Stockholm Abstract A novel method to account for dynamic speaker
Lecture 16 Speaker Recognition
Lecture 16 Speaker Recognition Information College, Shandong University @ Weihai Definition Method of recognizing a Person form his/her voice. Depends on Speaker Specific Characteristics To determine whether
Reinforcement Learning
Reinforcement Learning based Dialog Manager Speech Group Department of Signal Processing and Acoustics Katri Leino User Interface Group Department of Communications and Networking Aalto University, School
The Breath Segment in Expressive Speech
Computational Linguistics and Chinese Language Processing Vol. 12, No. 1, March 2007, pp. 17-32 17 The Association for Computational Linguistics and Chinese Language Processing The Breath Segment in Expressive
Intelligent Tutoring Systems using Reinforcement Learning to teach Autistic Students
Intelligent Tutoring Systems using Reinforcement Learning to teach Autistic Students B. H. Sreenivasa Sarma 1 and B. Ravindran 2 Department of Computer Science and Engineering, Indian Institute of Technology
Unsupervised Phoneme Segmentation in Continuous Speech
Unsupervised Phoneme Segmentation in Continuous Speech Stephanie Antetomaso Wheaton College Norton, MA USA antetomaso stephanie@wheatoncollege.edu Abstract A phonemic representation of speech is necessary
COMP150 DR Final Project Proposal
COMP150 DR Final Project Proposal Ari Brown and Julie Jiang October 26, 2017 Abstract The problem of sound classification has been studied in depth and has multiple applications related to identity discrimination,
Isolated Speech Recognition Using MFCC and DTW
Isolated Speech Recognition Using MFCC and DTW P.P.S.Subhashini Associate Professor, RVR & JC College of Engineering. ABSTRACT This paper describes an approach of isolated speech recognition by using the
Exploiting speaker segmentations for automatic role detection. An application to broadcast news documents.
Exploiting speaker segmentations for automatic role detection. An application to broadcast news documents. Benjamin Bigot Isabelle Ferrané IRIT - Université de Toulouse 118, route de Narbonne - 31062 Toulouse
Sensing and Modeling Human Networks using the Sociometer
Sensing and Modeling Human Networks using the Sociometer Tanzeem Choudhury and Alex Pentland Human Design Group 20 Ames Street, E15-384C Cambridge, MA02139 USA +1 617 253 0370 tanzeem@media.mit.edu Abstract
BUILDING COMPACT N-GRAM LANGUAGE MODELS INCREMENTALLY
BUILDING COMPACT N-GRAM LANGUAGE MODELS INCREMENTALLY Vesa Siivola Neural Networks Research Centre, Helsinki University of Technology, Finland Abstract In traditional n-gram language modeling, we collect
Advances in Music Information Retrieval using Deep Learning Techniques - Sid Pramod
Advances in Music Information Retrieval using Deep Learning Techniques - Sid Pramod Music Information Retrieval (MIR) Science of retrieving information from music. Includes tasks such as Query by Example,
Pass Phrase Based Speaker Recognition for Authentication
Pass Phrase Based Speaker Recognition for Authentication Heinz Hertlein, Dr. Robert Frischholz, Dr. Elmar Nöth* HumanScan GmbH Wetterkreuz 19a 91058 Erlangen/Tennenlohe, Germany * Chair for Pattern Recognition,
On-line recognition of handwritten characters
Chapter 8 On-line recognition of handwritten characters Vuokko Vuori, Matti Aksela, Ramūnas Girdziušas, Jorma Laaksonen, Erkki Oja 105 106 On-line recognition of handwritten characters 8.1 Introduction
A Low-Complexity Speaker-and-Word Recognition Application for Resource- Constrained Devices
A Low-Complexity Speaker-and-Word Application for Resource- Constrained Devices G. R. Dhinesh, G. R. Jagadeesh, T. Srikanthan Centre for High Performance Embedded Systems Nanyang Technological University,
A study of speaker adaptation for DNN-based speech synthesis
A study of speaker adaptation for DNN-based speech synthesis Zhizheng Wu, Pawel Swietojanski, Christophe Veaux, Steve Renals, Simon King The Centre for Speech Technology Research (CSTR) University of Edinburgh,
arxiv: v1 [cs.cl] 2 Jun 2015
Learning Speech Rate in Speech Recognition Xiangyu Zeng 1,3, Shi Yin 1,4, Dong Wang 1,2 1 CSLT, RIIT, Tsinghua University 2 TNList, Tsinghua University 3 Beijing University of Posts and Telecommunications
in 82 Dutch speakers. All of them were prompted to pronounce 10 sentences in four dierent languages : Dutch, English, French, and German. All the sent
MULTILINGUAL TEXT-INDEPENDENT SPEAKER IDENTIFICATION Georey Durou Faculte Polytechnique de Mons TCTS 31, Bld. Dolez B-7000 Mons, Belgium Email: durou@tcts.fpms.ac.be ABSTRACT In this paper, we investigate
Speech Emotion Recognition Using Deep Neural Network and Extreme. learning machine
INTERSPEECH 2014 Speech Emotion Recognition Using Deep Neural Network and Extreme Learning Machine Kun Han 1, Dong Yu 2, Ivan Tashev 2 1 Department of Computer Science and Engineering, The Ohio State University,
Gender Classification by Speech Analysis
Gender Classification by Speech Analysis BhagyaLaxmi Jena 1, Abhishek Majhi 2, Beda Prakash Panigrahi 3 1 Asst. Professor, Electronics & Tele-communication Dept., Silicon Institute of Technology 2,3 Students
BUILDING A STATISTICAL MODEL OF THE VOWEL SPACE FOR PHONETICIANS
BUILDING A STATISTICAL MODEL OF THE VOWEL SPACE FOR PHONETICIANS Matthew Aylett Human Communication Research Centre, University of Edinburgh email: matthewa@cogsci.ed.ac.uk ABSTRACT Vowel space data (A
Speech Synthesis. Tokyo Institute of Technology Department of fcomputer Science
Speech Synthesis Sadaoki Furui Tokyo Institute of Technology Department of fcomputer Science furui@cs.titech.ac.jp 0107-14 Pronouncing Acoustic dictionary segments and rules dictionary Text input Pronounce
A Survey on Hoeffding Tree Stream Data Classification Algorithms
CPUH-Research Journal: 2015, 1(2), 28-32 ISSN (Online): 2455-6076 http://www.cpuh.in/academics/academic_journals.php A Survey on Hoeffding Tree Stream Data Classification Algorithms Arvind Kumar 1*, Parminder
Speech Communication, Spring 2006
Speech Communication, Spring 2006 Lecture 3: Speech Coding and Synthesis Zheng-Hua Tan Department of Communication Technology Aalborg University, Denmark zt@kom.aau.dk Speech Communication, III, Zheng-Hua
Secondary Masters in Machine Learning
Secondary Masters in Machine Learning Student Handbook Revised 8/20/14 Page 1 Table of Contents Introduction... 3 Program Requirements... 4 Core Courses:... 5 Electives:... 6 Double Counting Courses:...
Speaker Indexing Using Neural Network Clustering of Vowel Spectra
International Journal of Speech Technology 1,143-149 (1997) @ 1997 Kluwer Academic Publishers. Manufactured in The Netherlands. Speaker Indexing Using Neural Network Clustering of Vowel Spectra DEB K.
Digital Speech Processing. Professor Lawrence Rabiner UCSB Dept. of Electrical and Computer Engineering Jan-March 2012
Digital Speech Processing Professor Lawrence Rabiner UCSB Dept. of Electrical and Computer Engineering Jan-March 2012 1 Course Description This course covers the basic principles of digital speech processing:
Hidden Markov Model-based speech synthesis
Hidden Markov Model-based speech synthesis Junichi Yamagishi, Korin Richmond, Simon King and many others Centre for Speech Technology Research University of Edinburgh, UK www.cstr.ed.ac.uk Note I did not
A NEW SPEAKER VERIFICATION APPROACH FOR BIOMETRIC SYSTEM
A NEW SPEAKER VERIFICATION APPROACH FOR BIOMETRIC SYSTEM J.INDRA 1 N.KASTHURI 2 M.BALASHANKAR 3 S.GEETHA MANJURI 4 1 Assistant Professor (Sl.G),Dept of Electronics and Instrumentation Engineering, 2 Professor,
Bidirectional LSTM Networks for Improved Phoneme Classification and Recognition
Bidirectional LSTM Networks for Improved Phoneme Classification and Recognition Alex Graves 1, Santiago Fernández 1, Jürgen Schmidhuber 1,2 1 IDSIA, Galleria 2, 6928 Manno-Lugano, Switzerland {alex,santiago,juergen}@idsia.ch
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May ISSN
International Journal of Scientific & Engineering Research, Volume 5, Issue 5, May-2014 600 Extraction of Prosodic Features for Speaker Recognition Technology and Voice Spectrum Analysis Authors: Nilu
Analysis of Importance of the prosodic Features for Automatic Sentence Modality Recognition in French in real Conditions
Analysis of Importance of the prosodic Features for Automatic Sentence Modality Recognition in French in real Conditions PAVEL KRÁL 1, JANA KLEČKOVÁ 1, CHRISTOPHE CERISARA 2 1 Dept. Informatics & Computer
Bayesian Deep Learning for Integrated Intelligence: Bridging the Gap between Perception and Inference
1 Bayesian Deep Learning for Integrated Intelligence: Bridging the Gap between Perception and Inference Hao Wang Department of Computer Science and Engineering Joint work with Naiyan Wang, Xingjian Shi,
Language, Mind, and Brain: Experience Alters perception
Language, Mind, and Brain: Experience Alters perception Chapter 8 The New Cognitive Neurosciences M. Gazzaniga (ed.) Sep 7, 2001 Relevant points from Stein et al. (Chap. 5) AES functions as an association
ELEC9723 Speech Processing
ELEC9723 Speech Processing COURSE INTRODUCTION Session 1, 2013 s Course Staff Course conveners: Dr. Vidhyasaharan Sethu, v.sethu@unsw.edu.au (EE304) Laboratory demonstrator: Nicholas Cummins, n.p.cummins@unsw.edu.au
Big Data Analytics Clustering and Classification
E6893 Big Data Analytics Lecture 4: Big Data Analytics Clustering and Classification Ching-Yung Lin, Ph.D. Adjunct Professor, Dept. of Electrical Engineering and Computer Science September 28th, 2017 1
Alberto Abad and Isabel Trancoso. L 2 F - Spoken Language Systems Lab INESC-ID / IST, Lisboa, Portugal
THE L 2 F LANGUAGE VERIFICATION SYSTEMS FOR ALBAYZIN-08 EVALUATION Alberto Abad and Isabel Trancoso L 2 F - Spoken Language Systems Lab INESC-ID / IST, Lisboa, Portugal {Alberto.Abad,Isabel.Trancoso}@l2f.inesc-id.pt
A Method for Translation of Paralinguistic Information
A Method for Translation of Paralinguistic Information Takatomo Kano, Sakriani Sakti, Shinnosuke Takamichi, Graham Neubig, Tomoki Toda, Satoshi Nakamura Graduate School of Information Science, Nara Institute
19 th INTERNATIONAL CONGRESS ON ACOUSTICS MADRID, 2-7 SEPTEMBER 2007
19 th INTERNATIONAL CONGRESS ON ACOUSTICS MADRID, 2-7 SEPTEMBER 2007 THE INFLUENCE OF LINGUISTIC AND EXTRA-LINGUISTIC INFORMATION ON SYNTHETIC SPEECH INTELLIGIBILITY PACS: 43.71 Bp Gardzielewska, Hanna
Auditory Context Recognition Using SVMs
Auditory Context Recognition Using SVMs Mikko Perttunen 1, Max Van Kleek 2, Ora Lassila 3, Jukka Riekki 1 1 Department of Electrical and Information Engineering, 90014 University of Oulu, Finland {first.last}@ee.oulu.fi
Prognostics and Health Management Approaches based on belief functions
Prognostics and Health Management Approaches based on belief functions FEMTO-ST institute / Dep. of Automation and Micromechatronics systems (AS2M), Besançon Emmanuel Ramasso Collaborated work with Dr.
Natural Speech Synthesizer for Blind Persons Using Hybrid Approach
Procedia Computer Science Volume 41, 2014, Pages 83 88 BICA 2014. 5th Annual International Conference on Biologically Inspired Cognitive Architectures Natural Speech Synthesizer for Blind Persons Using
Modelling the Emergence of Speech Sound Categories in Evolving Connectionist Systems. John Taylor Nikola Kasabov Richard Kilgour
DUNEDIN NEW ZEALAND Modelling the Emergence of Speech Sound Categories in Evolving Connectionist Systems John Taylor Nikola Kasabov Richard Kilgour The Information Science Discussion Paper Series Number
Speech Recognition at ICSI: Broadcast News and beyond
Speech Recognition at ICSI: Broadcast News and beyond Dan Ellis International Computer Science Institute, Berkeley CA Outline 1 2 3 The DARPA Broadcast News task Aspects of ICSI
Introduction to Deep Learning
Introduction to Deep Learning M S Ram Dept. of Computer Science & Engg. Indian Institute of Technology Kanpur Reading of Chap. 1 from Learning Deep Architectures for AI ; Yoshua Bengio; FTML Vol. 2, No.
Statistical Learning- Classification STAT 441/ 841, CM 764
Statistical Learning- Classification STAT 441/ 841, CM 764 Ali Ghodsi Department of Statistics and Actuarial Science University of Waterloo aghodsib@uwaterloo.ca Two Paradigms Classical Statistics Infer
Speech Accent Classification
Speech Accent Classification Corey Shih ctshih@stanford.edu 1. Introduction English is one of the most prevalent languages in the world, and is the one most commonly used for communication between native
SB2b Statistical Machine Learning Hilary Term 2017
SB2b Statistical Machine Learning Hilary Term 2017 Mihaela van der Schaar and Seth Flaxman Guest lecturer: Yee Whye Teh Department of Statistics Oxford Slides and other materials available at: http://www.oxford-man.ox.ac.uk/~mvanderschaar/home_
AL THE. The breakthrough machine learning platform for global speech recognition
AL THE The breakthrough machine learning platform for global speech recognition SEPTEMBER 2017 Introducing Speechmatics Automatic Linguist (AL) Automatic Speech Recognition (ASR) software has come a long
PUNJABI SPEECH SYNTHESIS SYSTEM USING HTK
PUNJABI SPEECH SYNTHESIS SYSTEM USING HTK Divya Bansal 1, Ankita Goel 2, Khushneet Jindal 3 School of Mathematics and Computer Applications, Thapar University, Patiala (Punjab) India 1 divyabansal150@yahoo.com