The use of speech-based solutions is an appealing alternative to communicate in human-robot interaction (HRI). An important challenge in this area is processing distant speech which is often noisy, and affected by reverberation and time-varying acoustic channels. It is important to investigate effective speech solutions, especially in dynamic environments where the robots and the users move, changing the distance and orientation between a speaker and the microphone. This paper addresses this problem in the context of speech emotion recognition (SER), which is an important task to understand the intention of the message and the underlying mental state of the user. We propose a novel setup with a PR2 robot that moves as target speech and ambient noise are simultaneously recorded. Our study not only analyzes the detrimental effect of distance speech in this dynamic robot-user setting for speech emotion recognition but also provides solutions to attenuate its effect. We evaluate the use of two beamforming schemes to spatially filter the speech signal using either delay-and-sum (D&S) or minimum variance distortionless response (MVDR). We consider the original training speech recorded in controlled situations, and simulated conditions where the training utterances are processed to simulate the target acoustic environment. We consider the case where the robot is moving (dynamic case) and not moving (static case). For speech emotion recognition, we explore two state-of-the-art classifiers using hand-crafted features implemented with the ladder network strategy and learned features implemented with the wav2vec 2.0 feature representation. MVDR led to a signal-to-noise ratio higher than the basic D&S method. However, both approaches provided very similar average concordance correlation coefficient (CCC) improvements equal to 116% with the HRI subsets using the ladder network trained with the original MSP-Podcast training utterances. For the wav2vec 2.0-based model, only D&S led to improvements. Surprisingly, the static and dynamic HRI testing subsets resulted in a similar average concordance correlation coefficient. Finally, simulating the acoustic environment in the training dataset provided the highest average concordance correlation coefficient scores with the HRI subsets that are just 29% and 22% lower than those obtained with the original training/testing utterances, with ladder network and wav2vec 2.0, respectively.

This research was funded by Conicyt-Chile under grants Fondecyt 1100195 and 11110391, and Team Research in Science and Technology ACT 1120.

This paper proposes a novel feature-space VTLN (vocal tract length normalization) method that models frequency warping as a linear interpolation of contiguous Mel filter-bank energies. The presented technique aims to reduce the distortion in the Mel filter-bank energy estimation due to the harmonic composition of voiced speech intervals and DFT (discrete Fourier transform) sampling when the central frequency of band-pass filters is shifted. This paper also proposes an analytical maximum likelihood (ML) method to estimate the optimal warping factor in the cepstral space. The presented interpolated filter-bank energy-based VTLN leads to relative reductions in WER (word error rate) as high as 11.2% and 7.6% when compared with the baseline system and standard VTLN, respectively, in a medium-vocabulary continuous speech recognition task. Also, the proposed VTLN scheme can provide significant reductions in WER when compared with state-of-the-art VTLN methods based on linear transforms in the cepstral feature-space. The warping factor estimated with the proposed VTLN approach shows more dependence on the speaker and more independence of the acoustic-phonetic content than the warping factor resulting from standard and state-of-the-art VTLN methods. Finally, the analytical ML-based optimization scheme presented here achieves almost the same reductions in WER as the ML grid search version of the technique with a computational load 20 times lower.

In this paper, a novel confidence-based reinforcement learning (RL) scheme to correct observation log-likelihoods and to address the problem of unsupervised compensation with limited estimation data is proposed. A two-step Viterbi decoding is presented which estimates a correction factor for the observation log-likelihoods that makes the recognized and neighboring HMMs more or less likely by using a confidence score. If regions in the output delivered by the recognizer exhibit low confidence scores, the second Viterbi decoding will tend to focus the search on neighboring models. In contrast, if recognized regions exhibit high confidence scores, the second Viterbi decoding will tend to retain the recognition output obtained at the first step. The proposed RL mechanism is modeled as the linear combination of two metrics or information sources: the acoustic model log-likelihood and the logarithm of a confidence metric. A criterion based on incremental conditional entropy maximization to optimize a linear combination of metrics or information sources online is also presented. The method requires only one utterance, as short as 0.7 s, and can lead to significant reductions in word error rate (WER) between 3% and 18%, depending on the task, training-testing conditions, and method used to optimize the proposed RL scheme. In contrast to ordinary feature compensation and model parameter adaptation methods, the confidence-based RL method takes place in the frame log-likelihood domain. Consequently, as shown in the results presented here, it is complementary to feature compensation and to model adaptation techniques.

This paper describes a two-step Viterbi decoding based on reinforcement learning and information theory with telephone speech.The idea is to strength or weaken HMM's by using Bayesbased confidence measure (BBCM) and distances between models.If HMM's in the N-best list show a low BBCM, the second Viterbi decoding will prioritize the search on neighboring models according to their distances to the N-best HMM's.The current reinforcement learning mechanism is modeled as the linear combination of two metrics or information sources.Moreover, a criterion based on incremental conditional entropy maximization to optimize a linear combination of metrics or information sources is also presented.As shown here, the method requires only one adapting utterance and can lead to a reduction in WER as high as 10.9%.

A novel framework based on Bayes-based confidence measure (BBCM) for Multiple Classifier System (MCS) fusion is proposed.As shown here, BBCM based MCS combination scheme corresponds to the ordinary Bayes fusion weighted by the reliability of each individual classifier.BBCM provides a formal model for heuristic weighting functions employed elsewhere.When compared with the ordinary Bayesian fusion, the proposed method leads to reductions as high as 20% and 50% in EER and the area below the ROC curve, respectively, in speaker verification.

This paper proposes a new paradigm to compensate for mismatch condition in speech recognition. A two-step Viterbi decoding based on reinforcement learning is described. The idea is to strength or weaken HMM’s by using Bayes-based confidence measure (BBCM) and distances between models. If HMM’s in the N-best list show a low BBCM, the second Viterbi decoding will prioritize the search on neighboring models according to their distances to the N-best HMM’s. As shown here, a reduction of 6% in WER is achieved in a task which results difficult for standard MAP and MLLR adaptation.

In this paper a compensation method is proposed to address the problem of limited enrolling data in speaker verification.Instead of adapting the client HMM, the technique presented here modifies the verification speech signals by maximizing the a posteriori p.d.f. in order to optimize the reduction in intra-speaker variability.The proposed approach can lead to reductions of 38.9% and 61.8% in EER and in the integral below the falseacceptation / false-rejection ROC curve, respectively.