Bhiksha Raj

Yidong Wang, Hao Chen, Yue Fan et al. · cmu, pku

Semi-supervised learning (SSL) improves model generalization by leveraging massive unlabeled data to augment limited labeled samples. However, currently, popular SSL evaluation protocols are often constrained to computer vision (CV) tasks. In addition, previous work typically trains deep neural networks from scratch, which is time-consuming and environmentally unfriendly. To address the above issues, we construct a Unified SSL Benchmark (USB) for classification by selecting 15 diverse, challenging, and comprehensive tasks from CV, natural language processing (NLP), and audio processing (Audio), on which we systematically evaluate the dominant SSL methods, and also open-source a modular and extensible codebase for fair evaluation of these SSL methods. We further provide the pre-trained versions of the state-of-the-art neural models for CV tasks to make the cost affordable for further tuning. USB enables the evaluation of a single SSL algorithm on more tasks from multiple domains but with less cost. Specifically, on a single NVIDIA V100, only 39 GPU days are required to evaluate FixMatch on 15 tasks in USB while 335 GPU days (279 GPU days on 4 CV datasets except for ImageNet) are needed on 5 CV tasks with TorchSSL.

Joseph Turian, Jordie Shier, Humair Raj Khan et al. · cmu

What audio embedding approach generalizes best to a wide range of downstream tasks across a variety of everyday domains without fine-tuning? The aim of the HEAR benchmark is to develop a general-purpose audio representation that provides a strong basis for learning in a wide variety of tasks and scenarios. HEAR evaluates audio representations using a benchmark suite across a variety of domains, including speech, environmental sound, and music. HEAR was launched as a NeurIPS 2021 shared challenge. In the spirit of shared exchange, each participant submitted an audio embedding model following a common API that is general-purpose, open-source, and freely available to use. Twenty-nine models by thirteen external teams were evaluated on nineteen diverse downstream tasks derived from sixteen datasets. Open evaluation code, submitted models and datasets are key contributions, enabling comprehensive and reproducible evaluation, as well as previously impossible longitudinal studies. It still remains an open question whether one single general-purpose audio representation can perform as holistically as the human ear.

Yidong Wang, Hao Chen, Qiang Heng et al. · cmu, pku

Semi-supervised Learning (SSL) has witnessed great success owing to the impressive performances brought by various methods based on pseudo labeling and consistency regularization. However, we argue that existing methods might fail to utilize the unlabeled data more effectively since they either use a pre-defined / fixed threshold or an ad-hoc threshold adjusting scheme, resulting in inferior performance and slow convergence. We first analyze a motivating example to obtain intuitions on the relationship between the desirable threshold and model's learning status. Based on the analysis, we hence propose FreeMatch to adjust the confidence threshold in a self-adaptive manner according to the model's learning status. We further introduce a self-adaptive class fairness regularization penalty to encourage the model for diverse predictions during the early training stage. Extensive experiments indicate the superiority of FreeMatch especially when the labeled data are extremely rare. FreeMatch achieves 5.78%, 13.59%, and 1.28% error rate reduction over the latest state-of-the-art method FlexMatch on CIFAR-10 with 1 label per class, STL-10 with 4 labels per class, and ImageNet with 100 labels per class, respectively. Moreover, FreeMatch can also boost the performance of imbalanced SSL. The codes can be found at https://github.com/microsoft/Semi-supervised-learning.

Kashu Yamazaki, Khoa Vo, Sang Truong et al. · cmu

Video paragraph captioning aims to generate a multi-sentence description of an untrimmed video with several temporal event locations in coherent storytelling. Following the human perception process, where the scene is effectively understood by decomposing it into visual (e.g. human, animal) and non-visual components (e.g. action, relations) under the mutual influence of vision and language, we first propose a visual-linguistic (VL) feature. In the proposed VL feature, the scene is modeled by three modalities including (i) a global visual environment; (ii) local visual main agents; (iii) linguistic scene elements. We then introduce an autoregressive Transformer-in-Transformer (TinT) to simultaneously capture the semantic coherence of intra- and inter-event contents within a video. Finally, we present a new VL contrastive loss function to guarantee learnt embedding features are matched with the captions semantics. Comprehensive experiments and extensive ablation studies on ActivityNet Captions and YouCookII datasets show that the proposed Visual-Linguistic Transformer-in-Transform (VLTinT) outperforms prior state-of-the-art methods on accuracy and diversity. Source code is made publicly available at: https://github.com/UARK-AICV/VLTinT.

Khoa Vo, Sang Truong, Kashu Yamazaki et al. · cmu

Temporal action proposal generation (TAPG) is a challenging task, which requires localizing action intervals in an untrimmed video. Intuitively, we as humans, perceive an action through the interactions between actors, relevant objects, and the surrounding environment. Despite the significant progress of TAPG, a vast majority of existing methods ignore the aforementioned principle of the human perceiving process by applying a backbone network into a given video as a black-box. In this paper, we propose to model these interactions with a multi-modal representation network, namely, Actors-Objects-Environment Interaction Network (AOE-Net). Our AOE-Net consists of two modules, i.e., perception-based multi-modal representation (PMR) and boundary-matching module (BMM). Additionally, we introduce adaptive attention mechanism (AAM) in PMR to focus only on main actors (or relevant objects) and model the relationships among them. PMR module represents each video snippet by a visual-linguistic feature, in which main actors and surrounding environment are represented by visual information, whereas relevant objects are depicted by linguistic features through an image-text model. BMM module processes the sequence of visual-linguistic features as its input and generates action proposals. Comprehensive experiments and extensive ablation studies on ActivityNet-1.3 and THUMOS-14 datasets show that our proposed AOE-Net outperforms previous state-of-the-art methods with remarkable performance and generalization for both TAPG and temporal action detection. To prove the robustness and effectiveness of AOE-Net, we further conduct an ablation study on egocentric videos, i.e. EPIC-KITCHENS 100 dataset. Source code is available upon acceptance.

Hao Chen, Ran Tao, Yue Fan et al. · cmu, pku

The critical challenge of Semi-Supervised Learning (SSL) is how to effectively leverage the limited labeled data and massive unlabeled data to improve the model's generalization performance. In this paper, we first revisit the popular pseudo-labeling methods via a unified sample weighting formulation and demonstrate the inherent quantity-quality trade-off problem of pseudo-labeling with thresholding, which may prohibit learning. To this end, we propose SoftMatch to overcome the trade-off by maintaining both high quantity and high quality of pseudo-labels during training, effectively exploiting the unlabeled data. We derive a truncated Gaussian function to weight samples based on their confidence, which can be viewed as a soft version of the confidence threshold. We further enhance the utilization of weakly-learned classes by proposing a uniform alignment approach. In experiments, SoftMatch shows substantial improvements across a wide variety of benchmarks, including image, text, and imbalanced classification.

Ankit Shah, Shuyi Chen, Kejun Zhou et al. · cmu

General-purpose embedding is highly desirable for few-shot even zero-shot learning in many application scenarios, including audio tasks. In order to understand representations better, we conducted a thorough error analysis and visualization of HEAR 2021 submission results. Inspired by the analysis, this work experiments with different front-end audio preprocessing methods, including Constant-Q Transform (CQT) and Short-time Fourier transform (STFT), and proposes a Batch Embedding Covariance Regularization (BECR) term to uncover a more holistic simulation of the frequency information received by the human auditory system. We tested the models on the suite of HEAR 2021 tasks, which encompass a broad category of tasks. Preliminary results show (1) the proposed BECR can incur a more dispersed embedding on the test set, (2) BECR improves the PaSST model without extra computation complexity, and (3) STFT preprocessing outperforms CQT in all tasks we tested. Github:https://github.com/ankitshah009/general_audio_embedding_hear_2021

Muqiao Yang, Joseph Konan, David Bick et al. · cmu

While deep learning based speech enhancement systems have made rapid progress in improving the quality of speech signals, they can still produce outputs that contain artifacts and can sound unnatural. We propose a novel approach to speech enhancement aimed at improving perceptual quality and naturalness of enhanced signals by optimizing for key characteristics of speech. We first identify key acoustic parameters that have been found to correlate well with voice quality (e.g. jitter, shimmer, and spectral flux) and then propose objective functions which are aimed at reducing the difference between clean speech and enhanced speech with respect to these features. The full set of acoustic features is the extended Geneva Acoustic Parameter Set (eGeMAPS), which includes 25 different attributes associated with perception of speech. Given the non-differentiable nature of these feature computation, we first build differentiable estimators of the eGeMAPS and then use them to fine-tune existing speech enhancement systems. Our approach is generic and can be applied to any existing deep learning based enhancement systems to further improve the enhanced speech signals. Experimental results conducted on the Deep Noise Suppression (DNS) Challenge dataset shows that our approach can improve the state-of-the-art deep learning based enhancement systems.

Yunyang Zeng, Joseph Konan, Shuo Han et al. · cmu

Speech enhancement models have greatly progressed in recent years, but still show limits in perceptual quality of their speech outputs. We propose an objective for perceptual quality based on temporal acoustic parameters. These are fundamental speech features that play an essential role in various applications, including speaker recognition and paralinguistic analysis. We provide a differentiable estimator for four categories of low-level acoustic descriptors involving: frequency-related parameters, energy or amplitude-related parameters, spectral balance parameters, and temporal features. Unlike prior work that looks at aggregated acoustic parameters or a few categories of acoustic parameters, our temporal acoustic parameter (TAP) loss enables auxiliary optimization and improvement of many fine-grain speech characteristics in enhancement workflows. We show that adding TAPLoss as an auxiliary objective in speech enhancement produces speech with improved perceptual quality and intelligibility. We use data from the Deep Noise Suppression 2020 Challenge to demonstrate that both time-domain models and time-frequency domain models can benefit from our method.

Roshan Sharma, Kenneth Zheng, Siddhant Arora et al. · cmu, meta-ai

End-to-end speech summarization has been shown to improve performance over cascade baselines. However, such models are difficult to train on very large inputs (dozens of minutes or hours) owing to compute restrictions and are hence trained with truncated model inputs. Truncation leads to poorer models, and a solution to this problem rests in block-wise modeling, i.e., processing a portion of the input frames at a time. In this paper, we develop a method that allows one to train summarization models on very long sequences in an incremental manner. Speech summarization is realized as a streaming process, where hypothesis summaries are updated every block based on new acoustic information. We devise and test strategies to pass semantic context across the blocks. Experiments on the How2 dataset demonstrate that the proposed block-wise training method improves by 3 points absolute on ROUGE-L over a truncated input baseline.

Raphael Olivier, Bhiksha Raj

Whisper is a recent Automatic Speech Recognition (ASR) model displaying impressive robustness to both out-of-distribution inputs and random noise. In this work, we show that this robustness does not carry over to adversarial noise. We show that we can degrade Whisper performance dramatically, or even transcribe a target sentence of our choice, by generating very small input perturbations with Signal Noise Ratio of 35-45dB. We also show that by fooling the Whisper language detector we can very easily degrade the performance of multilingual models. These vulnerabilities of a widely popular open-source model have practical security implications and emphasize the need for adversarially robust ASR.

Muqiao Yang, Joseph Konan, David Bick et al. · cmu

Despite rapid advancement in recent years, current speech enhancement models often produce speech that differs in perceptual quality from real clean speech. We propose a learning objective that formalizes differences in perceptual quality, by using domain knowledge of acoustic-phonetics. We identify temporal acoustic parameters -- such as spectral tilt, spectral flux, shimmer, etc. -- that are non-differentiable, and we develop a neural network estimator that can accurately predict their time-series values across an utterance. We also model phoneme-specific weights for each feature, as the acoustic parameters are known to show different behavior in different phonemes. We can add this criterion as an auxiliary loss to any model that produces speech, to optimize speech outputs to match the values of clean speech in these features. Experimentally we show that it improves speech enhancement workflows in both time-domain and time-frequency domain, as measured by standard evaluation metrics. We also provide an analysis of phoneme-dependent improvement on acoustic parameters, demonstrating the additional interpretability that our method provides. This analysis can suggest which features are currently the bottleneck for improvement.

Kai Qiu, Xiang Li, Hao Chen et al.

Audio generation has achieved remarkable progress with the advance of sophisticated generative models, such as diffusion models (DMs) and autoregressive (AR) models. However, due to the naturally significant sequence length of audio, the efficiency of audio generation remains an essential issue to be addressed, especially for AR models that are incorporated in large language models (LLMs). In this paper, we analyze the token length of audio tokenization and propose a novel \textbf{S}cale-level \textbf{A}udio \textbf{T}okenizer (SAT), with improved residual quantization. Based on SAT, a scale-level \textbf{A}coustic \textbf{A}uto\textbf{R}egressive (AAR) modeling framework is further proposed, which shifts the next-token AR prediction to next-scale AR prediction, significantly reducing the training cost and inference time. To validate the effectiveness of the proposed approach, we comprehensively analyze design choices and demonstrate the proposed AAR framework achieves a remarkable \textbf{35}$\times$ faster inference speed and +\textbf{1.33} Fréchet Audio Distance (FAD) against baselines on the AudioSet benchmark. Code: \url{https://github.com/qiuk2/AAR}.

Hao Chen, Yue Fan, Yidong Wang et al. · pku

Semi-supervised learning (SSL) has shown great promise in leveraging unlabeled data to improve model performance. While standard SSL assumes uniform data distribution, we consider a more realistic and challenging setting called imbalanced SSL, where imbalanced class distributions occur in both labeled and unlabeled data. Although there are existing endeavors to tackle this challenge, their performance degenerates when facing severe imbalance since they can not reduce the class imbalance sufficiently and effectively. In this paper, we study a simple yet overlooked baseline -- SimiS -- which tackles data imbalance by simply supplementing labeled data with pseudo-labels, according to the difference in class distribution from the most frequent class. Such a simple baseline turns out to be highly effective in reducing class imbalance. It outperforms existing methods by a significant margin, e.g., 12.8%, 13.6%, and 16.7% over previous SOTA on CIFAR100-LT, FOOD101-LT, and ImageNet127 respectively. The reduced imbalance results in faster convergence and better pseudo-label accuracy of SimiS. The simplicity of our method also makes it possible to be combined with other re-balancing techniques to improve the performance further. Moreover, our method shows great robustness to a wide range of data distributions, which holds enormous potential in practice. Code will be publicly available.

Xiang Li, Yandong Wen, Muqiao Yang et al. · cmu

Previous works on voice-face matching and voice-guided face synthesis demonstrate strong correlations between voice and face, but mainly rely on coarse semantic cues such as gender, age, and emotion. In this paper, we aim to investigate the capability of reconstructing the 3D facial shape from voice from a geometry perspective without any semantic information. We propose a voice-anthropometric measurement (AM)-face paradigm, which identifies predictable facial AMs from the voice and uses them to guide 3D face reconstruction. By leveraging AMs as a proxy to link the voice and face geometry, we can eliminate the influence of unpredictable AMs and make the face geometry tractable. Our approach is evaluated on our proposed dataset with ground-truth 3D face scans and corresponding voice recordings, and we find significant correlations between voice and specific parts of the face geometry, such as the nasal cavity and cranium. Our work offers a new perspective on voice-face correlation and can serve as a good empirical study for anthropometry science.

Roshan Sharma, Hira Dhamyal, Bhiksha Raj et al. · cmu, meta-ai

Traditionally, in paralinguistic analysis for emotion detection from speech, emotions have been identified with discrete or dimensional (continuous-valued) labels. Accordingly, models that have been proposed for emotion detection use one or the other of these label types. However, psychologists like Russell and Plutchik have proposed theories and models that unite these views, maintaining that these representations have shared and complementary information. This paper is an attempt to validate these viewpoints computationally. To this end, we propose a model to jointly predict continuous and discrete emotional attributes and show how the relationship between these can be utilized to improve the robustness and performance of emotion recognition tasks. Our approach comprises multi-task and hierarchical multi-task learning frameworks that jointly model the relationships between continuous-valued and discrete emotion labels. Experimental results on two widely used datasets (IEMOCAP and MSPPodcast) for speech-based emotion recognition show that our model results in statistically significant improvements in performance over strong baselines with non-unified approaches. We also demonstrate that using one type of label (discrete or continuous-valued) for training improves recognition performance in tasks that use the other type of label. Experimental results and reasoning for this approach (called the mismatched training approach) are also presented.

Roshan Sharma, Tyler Vuong, Mark Lindsey et al. · cmu, meta-ai

This work presents a multitask approach to the simultaneous estimation of age, country of origin, and emotion given vocal burst audio for the 2022 ICML Expressive Vocalizations Challenge ExVo-MultiTask track. The method of choice utilized a combination of spectro-temporal modulation and self-supervised features, followed by an encoder-decoder network organized in a multitask paradigm. We evaluate the complementarity between the tasks posed by examining independent task-specific and joint models, and explore the relative strengths of different feature sets. We also introduce a simple score fusion mechanism to leverage the complementarity of different feature sets for this task. We find that robust data preprocessing in conjunction with score fusion over spectro-temporal receptive field and HuBERT models achieved our best ExVo-MultiTask test score of 0.412.

Roshan Sharma, Bhiksha Raj · cmu, meta-ai

Transformers are among the state of the art for many tasks in speech, vision, and natural language processing, among others. Self-attentions, which are crucial contributors to this performance have quadratic computational complexity, which makes training on longer input sequences challenging. Prior work has produced state-of-the-art transformer variants with linear attention, however, current models sacrifice performance to achieve efficient implementations. In this work, we develop a novel linear transformer by examining the properties of the key-query product within self-attentions. Our model outperforms state of the art approaches on speech recognition and speech summarization, resulting in 1 % absolute WER improvement on the Librispeech-100 speech recognition benchmark and a new INTERVIEW speech recognition benchmark, and 5 points on ROUGE for summarization with How2.

Joseph Konan, Ojas Bhargave, Shikhar Agnihotri et al. · cmu

In this paper, we present a method for fine-tuning models trained on the Deep Noise Suppression (DNS) 2020 Challenge to improve their performance on Voice over Internet Protocol (VoIP) applications. Our approach involves adapting the DNS 2020 models to the specific acoustic characteristics of VoIP communications, which includes distortion and artifacts caused by compression, transmission, and platform-specific processing. To this end, we propose a multi-task learning framework for VoIP-DNS that jointly optimizes noise suppression and VoIP-specific acoustics for speech enhancement. We evaluate our approach on a diverse VoIP scenarios and show that it outperforms both industry performance and state-of-the-art methods for speech enhancement on VoIP applications. Our results demonstrate the potential of models trained on DNS-2020 to be improved and tailored to different VoIP platforms using VoIP-DNS, whose findings have important applications in areas such as speech recognition, voice assistants, and telecommunication.

Xiang Li, Haoyuan Cao, Shijie Zhao et al. · pku

Video salient object detection (VSOD), as a fundamental computer vision problem, has been extensively discussed in the last decade. However, all existing works focus on addressing the VSOD problem in 2D scenarios. With the rapid development of VR devices, panoramic videos have been a promising alternative to 2D videos to provide immersive feelings of the real world. In this paper, we aim to tackle the video salient object detection problem for panoramic videos, with their corresponding ambisonic audios. A multimodal fusion module equipped with two pseudo-siamese audio-visual context fusion (ACF) blocks is proposed to effectively conduct audio-visual interaction. The ACF block equipped with spherical positional encoding enables the fusion in the 3D context to capture the spatial correspondence between pixels and sound sources from the equirectangular frames and ambisonic audios. Experimental results verify the effectiveness of our proposed components and demonstrate that our method achieves state-of-the-art performance on the ASOD60K dataset.

Xiang Li, Jinglu Wang, Xiaohao Xu et al.

Audiovisual segmentation (AVS) is a challenging task that aims to segment visual objects in videos according to their associated acoustic cues. With multiple sound sources and background disturbances involved, establishing robust correspondences between audio and visual contents poses unique challenges due to (1) complex entanglement across sound sources and (2) frequent changes in the occurrence of distinct sound events. Assuming sound events occur independently, the multi-source semantic space can be represented as the Cartesian product of single-source sub-spaces. We are motivated to decompose the multi-source audio semantics into single-source semantics for more effective interactions with visual content. We propose a semantic decomposition method based on product quantization, where the multi-source semantics can be decomposed and represented by several disentangled and noise-suppressed single-source semantics. Furthermore, we introduce a global-to-local quantization mechanism, which distills knowledge from stable global (clip-level) features into local (frame-level) ones, to handle frequent changes in audio semantics. Extensive experiments demonstrate that our semantically decomposed audio representation significantly improves AVS performance, e.g., +21.2% mIoU on the challenging AVS-Semantic benchmark with ResNet50 backbone. https://github.com/lxa9867/QSD.

Hao Chen, Jindong Wang, Ankit Shah et al.

Pre-training on large-scale datasets and then fine-tuning on downstream tasks have become a standard practice in deep learning. However, pre-training data often contain label noise that may adversely affect the generalization of the model. This paper aims to understand the nature of noise in pre-training datasets and to mitigate its impact on downstream tasks. More specifically, through extensive experiments of supervised pre-training models on synthetic noisy ImageNet-1K and YFCC15M datasets, we demonstrate that while slight noise in pre-training can benefit in-domain (ID) transfer performance, where the training and testing data share the same distribution, it always deteriorates out-of-domain (OOD) performance, where training and testing data distribution are different. We empirically verify that the reason behind is noise in pre-training shapes the feature space differently. We then propose a light-weight black-box tuning method (NMTune) to affine the feature space to mitigate the malignant effect of noise and improve generalization on both ID and OOD tasks, considering one may not be able to fully fine-tune or even access the pre-trained models. We conduct practical experiments on popular vision and language models that are pre-trained on noisy data for evaluation of our approach. Our analysis and results show the importance of this interesting and novel research direction, which we term Noisy Model Learning.

Larry Tang, Po Hao Chou, Yi Yu Zheng et al. · cmu

Ontologies encompass a formal representation of knowledge through the definition of concepts or properties of a domain, and the relationships between those concepts. In this work, we seek to investigate whether using this ontological information will improve learning from weakly labeled data, which are easier to collect since it requires only the presence or absence of an event to be known. We use the AudioSet ontology and dataset, which contains audio clips weakly labeled with the ontology concepts and the ontology providing the "Is A" relations between the concepts. We first re-implemented the model proposed by soundevent_ontology with modification to fit the multi-label scenario and then expand on that idea by using a Graph Convolutional Network (GCN) to model the ontology information to learn the concepts. We find that the baseline Siamese does not perform better by incorporating ontology information in the weak and multi-label scenario, but that the GCN does capture the ontology knowledge better for weak, multi-labeled data. In our experiments, we also investigate how different modules can tolerate noises introduced from weak labels and better incorporate ontology information. Our best Siamese-GCN model achieves mAP=0.45 and AUC=0.87 for lower-level concepts and mAP=0.72 and AUC=0.86 for higher-level concepts, which is an improvement over the baseline Siamese but about the same as our models that do not use ontology information.

Xiang Li, Jinglu Wang, Xiaohao Xu et al.

Referring Video Object Segmentation (R-VOS) is a challenging task that aims to segment an object in a video based on a linguistic expression. Most existing R-VOS methods have a critical assumption: the object referred to must appear in the video. This assumption, which we refer to as semantic consensus, is often violated in real-world scenarios, where the expression may be queried against false videos. In this work, we highlight the need for a robust R-VOS model that can handle semantic mismatches. Accordingly, we propose an extended task called Robust R-VOS, which accepts unpaired video-text inputs. We tackle this problem by jointly modeling the primary R-VOS problem and its dual (text reconstruction). A structural text-to-text cycle constraint is introduced to discriminate semantic consensus between video-text pairs and impose it in positive pairs, thereby achieving multi-modal alignment from both positive and negative pairs. Our structural constraint effectively addresses the challenge posed by linguistic diversity, overcoming the limitations of previous methods that relied on the point-wise constraint. A new evaluation dataset, R\textsuperscript{2}-Youtube-VOSis constructed to measure the model robustness. Our model achieves state-of-the-art performance on R-VOS benchmarks, Ref-DAVIS17 and Ref-Youtube-VOS, and also our R\textsuperscript{2}-Youtube-VOS~dataset.

Thanh-Dat Truong, Ngan Le, Bhiksha Raj et al.

Although Domain Adaptation in Semantic Scene Segmentation has shown impressive improvement in recent years, the fairness concerns in the domain adaptation have yet to be well defined and addressed. In addition, fairness is one of the most critical aspects when deploying the segmentation models into human-related real-world applications, e.g., autonomous driving, as any unfair predictions could influence human safety. In this paper, we propose a novel Fairness Domain Adaptation (FREDOM) approach to semantic scene segmentation. In particular, from the proposed formulated fairness objective, a new adaptation framework will be introduced based on the fair treatment of class distributions. Moreover, to generally model the context of structural dependency, a new conditional structural constraint is introduced to impose the consistency of predicted segmentation. Thanks to the proposed Conditional Structure Network, the self-attention mechanism has sufficiently modeled the structural information of segmentation. Through the ablation studies, the proposed method has shown the performance improvement of the segmentation models and promoted fairness in the model predictions. The experimental results on the two standard benchmarks, i.e., SYNTHIA $\to$ Cityscapes and GTA5 $\to$ Cityscapes, have shown that our method achieved State-of-the-Art (SOTA) performance.

Muqiao Yang, Chunlei Zhang, Yong Xu et al. · cmu

Speech enhancement aims to improve the quality of speech signals in terms of quality and intelligibility, and speech editing refers to the process of editing the speech according to specific user needs. In this paper, we propose a Unified Speech Enhancement and Editing (uSee) model with conditional diffusion models to handle various tasks at the same time in a generative manner. Specifically, by providing multiple types of conditions including self-supervised learning embeddings and proper text prompts to the score-based diffusion model, we can enable controllable generation of the unified speech enhancement and editing model to perform corresponding actions on the source speech. Our experiments show that our proposed uSee model can achieve superior performance in both speech denoising and dereverberation compared to other related generative speech enhancement models, and can perform speech editing given desired environmental sound text description, signal-to-noise ratios (SNR), and room impulse responses (RIR). Demos of the generated speech are available at https://muqiaoy.github.io/usee.

Umberto Cappellazzo, Enrico Fini, Muqiao Yang et al. · cmu

Recently, neural networks have shown impressive progress across diverse fields, with speech processing being no exception. However, recent breakthroughs in this area require extensive offline training using large datasets and tremendous computing resources. Unfortunately, these models struggle to retain their previously acquired knowledge when learning new tasks continually, and retraining from scratch is almost always impractical. In this paper, we investigate the problem of learning sequence-to-sequence models for spoken language understanding in a class-incremental learning (CIL) setting and we propose COCONUT, a CIL method that relies on the combination of experience replay and contrastive learning. Through a modified version of the standard supervised contrastive loss applied only to the rehearsal samples, COCONUT preserves the learned representations by pulling closer samples from the same class and pushing away the others. Moreover, we leverage a multimodal contrastive loss that helps the model learn more discriminative representations of the new data by aligning audio and text features. We also investigate different contrastive designs to combine the strengths of the contrastive loss with teacher-student architectures used for distillation. Experiments on two established SLU datasets reveal the effectiveness of our proposed approach and significant improvements over the baselines. We also show that COCONUT can be combined with methods that operate on the decoder side of the model, resulting in further metrics improvements.

Hira Dhamyal, Benjamin Elizalde, Soham Deshmukh et al.

Emotions lie on a broad continuum and treating emotions as a discrete number of classes limits the ability of a model to capture the nuances in the continuum. The challenge is how to describe the nuances of emotions and how to enable a model to learn the descriptions. In this work, we devise a method to automatically create a description (or prompt) for a given audio by computing acoustic properties, such as pitch, loudness, speech rate, and articulation rate. We pair a prompt with its corresponding audio using 5 different emotion datasets. We trained a neural network model using these audio-text pairs. Then, we evaluate the model using one more dataset. We investigate how the model can learn to associate the audio with the descriptions, resulting in performance improvement of Speech Emotion Recognition and Speech Audio Retrieval. We expect our findings to motivate research describing the broad continuum of emotion

Raphael Olivier, Bhiksha Raj

Robustness to adversarial attacks is typically evaluated with adversarial accuracy. While essential, this metric does not capture all aspects of robustness and in particular leaves out the question of how many perturbations can be found for each point. In this work, we introduce an alternative approach, adversarial sparsity, which quantifies how difficult it is to find a successful perturbation given both an input point and a constraint on the direction of the perturbation. We show that sparsity provides valuable insight into neural networks in multiple ways: for instance, it illustrates important differences between current state-of-the-art robust models them that accuracy analysis does not, and suggests approaches for improving their robustness. When applying broken defenses effective against weak attacks but not strong ones, sparsity can discriminate between the totally ineffective and the partially effective defenses. Finally, with sparsity we can measure increases in robustness that do not affect accuracy: we show for example that data augmentation can by itself increase adversarial robustness, without using adversarial training.

Roshan Sharma, Suwon Shon, Mark Lindsey et al.

Reference summaries for abstractive speech summarization require human annotation, which can be performed by listening to an audio recording or by reading textual transcripts of the recording. In this paper, we examine whether summaries based on annotators listening to the recordings differ from those based on annotators reading transcripts. Using existing intrinsic evaluation based on human evaluation, automatic metrics, LLM-based evaluation, and a retrieval-based reference-free method. We find that summaries are indeed different based on the source modality, and that speech-based summaries are more factually consistent and information-selective than transcript-based summaries. Meanwhile, transcript-based summaries are impacted by recognition errors in the source, and expert-written summaries are more informative and reliable. We make all the collected data and analysis code public(https://github.com/cmu-mlsp/interview_humanssum) to facilitate the reproduction of our work and advance research in this area.

Joseph Konan, Shikhar Agnihotri, Ojas Bhargave et al. · cmu

Within the ambit of VoIP (Voice over Internet Protocol) telecommunications, the complexities introduced by acoustic transformations merit rigorous analysis. This research, rooted in the exploration of proprietary sender-side denoising effects, meticulously evaluates platforms such as Google Meets and Zoom. The study draws upon the Deep Noise Suppression (DNS) 2020 dataset, ensuring a structured examination tailored to various denoising settings and receiver interfaces. A methodological novelty is introduced via Blinder-Oaxaca decomposition, traditionally an econometric tool, repurposed herein to analyze acoustic-phonetic perturbations within VoIP systems. To further ground the implications of these transformations, psychoacoustic metrics, specifically PESQ and STOI, were used to explain of perceptual quality and intelligibility. Cumulatively, the insights garnered underscore the intricate landscape of VoIP-influenced acoustic dynamics. In addition to the primary findings, a multitude of metrics are reported, extending the research purview. Moreover, out-of-domain benchmarking for both time and time-frequency domain speech enhancement models is included, thereby enhancing the depth and applicability of this inquiry.

Ankit Shah, Fuyu Tang, Zelin Ye et al. · cmu

Weak-label learning is a challenging task that requires learning from data "bags" containing positive and negative instances, but only the bag labels are known. The pool of negative instances is usually larger than positive instances, thus making selecting the most informative negative instance critical for performance. Such a selection strategy for negative instances from each bag is an open problem that has not been well studied for weak-label learning. In this paper, we study several sampling strategies that can measure the usefulness of negative instances for weak-label learning and select them accordingly. We test our method on CIFAR-10 and AudioSet datasets and show that it improves the weak-label classification performance and reduces the computational cost compared to random sampling methods. Our work reveals that negative instances are not all equally irrelevant, and selecting them wisely can benefit weak-label learning.

Muhammad A. Shah, Bhiksha Raj

Deep neural networks (DNNs) have been shown to be vulnerable to adversarial attacks -- subtle, perceptually indistinguishable perturbations of inputs that change the response of the model. In the context of vision, we hypothesize that an important contributor to the robustness of human visual perception is constant exposure to low-fidelity visual stimuli in our peripheral vision. To investigate this hypothesis, we develop \RBlur, an image transform that simulates the loss in fidelity of peripheral vision by blurring the image and reducing its color saturation based on the distance from a given fixation point. We show that compared to DNNs trained on the original images, DNNs trained on images transformed by \RBlur are substantially more robust to adversarial attacks, as well as other, non-adversarial, corruptions, achieving up to 25\% higher accuracy on perturbed data.

Liao Qu, Xianwei Zou, Xiang Li et al.

This work unveils the enigmatic link between phonemes and facial features. Traditional studies on voice-face correlations typically involve using a long period of voice input, including generating face images from voices and reconstructing 3D face meshes from voices. However, in situations like voice-based crimes, the available voice evidence may be short and limited. Additionally, from a physiological perspective, each segment of speech -- phoneme -- corresponds to different types of airflow and movements in the face. Therefore, it is advantageous to discover the hidden link between phonemes and face attributes. In this paper, we propose an analysis pipeline to help us explore the voice-face relationship in a fine-grained manner, i.e., phonemes v.s. facial anthropometric measurements (AM). We build an estimator for each phoneme-AM pair and evaluate the correlation through hypothesis testing. Our results indicate that AMs are more predictable from vowels compared to consonants, particularly with plosives. Additionally, we observe that if a specific AM exhibits more movement during phoneme pronunciation, it is more predictable. Our findings support those in physiology regarding correlation and lay the groundwork for future research on speech-face multimodal learning.

Xiang Li, Jinglu Wang, Xiaohao Xu et al.

Video instance segmentation (VIS) aims at classifying, segmenting and tracking object instances in video sequences. Recent transformer-based neural networks have demonstrated their powerful capability of modeling spatio-temporal correlations for the VIS task. Relying on video- or clip-level input, they suffer from high latency and computational cost. We propose a robust context fusion network to tackle VIS in an online fashion, which predicts instance segmentation frame-by-frame with a few preceding frames. To acquire the precise and temporal-consistent prediction for each frame efficiently, the key idea is to fuse effective and compact context from reference frames into the target frame. Considering the different effects of reference and target frames on the target prediction, we first summarize contextual features through importance-aware compression. A transformer encoder is adopted to fuse the compressed context. Then, we leverage an order-preserving instance embedding to convey the identity-aware information and correspond the identities to predicted instance masks. We demonstrate that our robust fusion network achieves the best performance among existing online VIS methods and is even better than previously published clip-level methods on the Youtube-VIS 2019 and 2021 benchmarks. In addition, visual objects often have acoustic signatures that are naturally synchronized with them in audio-bearing video recordings. By leveraging the flexibility of our context fusion network on multi-modal data, we further investigate the influence of audios on the video-dense prediction task, which has never been discussed in existing works. We build up an Audio-Visual Instance Segmentation dataset, and demonstrate that acoustic signals in the wild scenarios could benefit the VIS task.

Pha Nguyen, Kha Gia Quach, John Gauch et al.

Multiple Object Tracking (MOT) aims to find bounding boxes and identities of targeted objects in consecutive video frames. While fully-supervised MOT methods have achieved high accuracy on existing datasets, they cannot generalize well on a newly obtained dataset or a new unseen domain. In this work, we first address the MOT problem from the cross-domain point of view, imitating the process of new data acquisition in practice. Then, a new cross-domain MOT adaptation from existing datasets is proposed without any pre-defined human knowledge in understanding and modeling objects. It can also learn and update itself from the target data feedback. The intensive experiments are designed on four challenging settings, including MOTSynth to MOT17, MOT17 to MOT20, MOT17 to VisDrone, and MOT17 to DanceTrack. We then prove the adaptability of the proposed self-supervised learning strategy. The experiments also show superior performance on tracking metrics MOTA and IDF1, compared to fully supervised, unsupervised, and self-supervised state-of-the-art methods.

Shentong Mo, Jingfei Xia, Xiaoqing Tan et al.

Spatio-temporal action recognition has been a challenging task that involves detecting where and when actions occur. Current state-of-the-art action detectors are mostly anchor-based, requiring sensitive anchor designs and huge computations due to calculating large numbers of anchor boxes. Motivated by nascent anchor-free approaches, we propose Point3D, a flexible and computationally efficient network with high precision for spatio-temporal action recognition. Our Point3D consists of a Point Head for action localization and a 3D Head for action classification. Firstly, Point Head is used to track center points and knot key points of humans to localize the bounding box of an action. These location features are then piped into a time-wise attention to learn long-range dependencies across frames. The 3D Head is later deployed for the final action classification. Our Point3D achieves state-of-the-art performance on the JHMDB, UCF101-24, and AVA benchmarks in terms of frame-mAP and video-mAP. Comprehensive ablation studies also demonstrate the effectiveness of each module proposed in our Point3D.

Raphael Olivier, Hadi Abdullah, Bhiksha Raj

A targeted adversarial attack produces audio samples that can force an Automatic Speech Recognition (ASR) system to output attacker-chosen text. To exploit ASR models in real-world, black-box settings, an adversary can leverage the transferability property, i.e. that an adversarial sample produced for a proxy ASR can also fool a different remote ASR. However recent work has shown that transferability against large ASR models is very difficult. In this work, we show that modern ASR architectures, specifically ones based on Self-Supervised Learning, are in fact vulnerable to transferability. We successfully demonstrate this phenomenon by evaluating state-of-the-art self-supervised ASR models like Wav2Vec2, HuBERT, Data2Vec and WavLM. We show that with low-level additive noise achieving a 30dB Signal-Noise Ratio, we can achieve target transferability with up to 80% accuracy. Next, we 1) use an ablation study to show that Self-Supervised learning is the main cause of that phenomenon, and 2) we provide an explanation for this phenomenon. Through this we show that modern ASR architectures are uniquely vulnerable to adversarial security threats.

Shentong Mo, Bhiksha Raj

Audio-visual segmentation is a challenging task that aims to predict pixel-level masks for sound sources in a video. Previous work applied a comprehensive manually designed architecture with countless pixel-wise accurate masks as supervision. However, these pixel-level masks are expensive and not available in all cases. In this work, we aim to simplify the supervision as the instance-level annotation, i.e., weakly-supervised audio-visual segmentation. We present a novel Weakly-Supervised Audio-Visual Segmentation framework, namely WS-AVS, that can learn multi-scale audio-visual alignment with multi-scale multiple-instance contrastive learning for audio-visual segmentation. Extensive experiments on AVSBench demonstrate the effectiveness of our WS-AVS in the weakly-supervised audio-visual segmentation of single-source and multi-source scenarios.

Yutian Chen, Hao Kang, Vivian Zhai et al.

This paper introduces a novel approach for identifying the possible large language models (LLMs) involved in text generation. Instead of adding an additional classification layer to a base LM, we reframe the classification task as a next-token prediction task and directly fine-tune the base LM to perform it. We utilize the Text-to-Text Transfer Transformer (T5) model as the backbone for our experiments. We compared our approach to the more direct approach of utilizing hidden states for classification. Evaluation shows the exceptional performance of our method in the text classification task, highlighting its simplicity and efficiency. Furthermore, interpretability studies on the features extracted by our model reveal its ability to differentiate distinctive writing styles among various LLMs even in the absence of an explicit classifier. We also collected a dataset named OpenLLMText, containing approximately 340k text samples from human and LLMs, including GPT3.5, PaLM, LLaMA, and GPT2.

Muhammad Ahmed Shah, Bhiksha Raj

The vulnerability to adversarial perturbations is a major flaw of Deep Neural Networks (DNNs) that raises question about their reliability when in real-world scenarios. On the other hand, human perception, which DNNs are supposed to emulate, is highly robust to such perturbations, indicating that there may be certain features of the human perception that make it robust but are not represented in the current class of DNNs. One such feature is that the activity of biological neurons is correlated and the structure of this correlation tends to be rather rigid over long spans of times, even if it hampers performance and learning. We hypothesize that integrating such constraints on the activations of a DNN would improve its adversarial robustness, and, to test this hypothesis, we have developed the Self-Consistent Activation (SCA) layer, which comprises of neurons whose activations are consistent with each other, as they conform to a fixed, but learned, covariability pattern. When evaluated on image and sound recognition tasks, the models with a SCA layer achieved high accuracy, and exhibited significantly greater robustness than multi-layer perceptron models to state-of-the-art Auto-PGD adversarial attacks \textit{without being trained on adversarially perturbed data

Samiran Gode, Supreeth Bare, Bhiksha Raj et al.

Our paper aims to analyze political polarization in US political system using Language Models, and thereby help candidates make an informed decision. The availability of this information will help voters understand their candidates views on the economy, healthcare, education and other social issues. Our main contributions are a dataset extracted from Wikipedia that spans the past 120 years and a Language model based method that helps analyze how polarized a candidate is. Our data is divided into 2 parts, background information and political information about a candidate, since our hypothesis is that the political views of a candidate should be based on reason and be independent of factors such as birthplace, alma mater, etc. We further split this data into 4 phases chronologically, to help understand if and how the polarization amongst candidates changes. This data has been cleaned to remove biases. To understand the polarization we begin by showing results from some classical language models in Word2Vec and Doc2Vec. And then use more powerful techniques like the Longformer, a transformer based encoder, to assimilate more information and find the nearest neighbors of each candidate based on their political view and their background.

Chonghan Chen, Qi Jiang, Chih-Hao Wang et al.

Visual grounding is a task that aims to locate a target object according to a natural language expression. As a multi-modal task, feature interaction between textual and visual inputs is vital. However, previous solutions mainly handle each modality independently before fusing them together, which does not take full advantage of relevant textual information while extracting visual features. To better leverage the textual-visual relationship in visual grounding, we propose a Query-conditioned Convolution Module (QCM) that extracts query-aware visual features by incorporating query information into the generation of convolutional kernels. With our proposed QCM, the downstream fusion module receives visual features that are more discriminative and focused on the desired object described in the expression, leading to more accurate predictions. Extensive experiments on three popular visual grounding datasets demonstrate that our method achieves state-of-the-art performance. In addition, the query-aware visual features are informative enough to achieve comparable performance to the latest methods when directly used for prediction without further multi-modal fusion.

Muhammad Ahmed Shah, Roshan Sharma, Hira Dhamyal et al.

It has been shown that Large Language Model (LLM) alignments can be circumvented by appending specially crafted attack suffixes with harmful queries to elicit harmful responses. To conduct attacks against private target models whose characterization is unknown, public models can be used as proxies to fashion the attack, with successful attacks being transferred from public proxies to private target models. The success rate of attack depends on how closely the proxy model approximates the private model. We hypothesize that for attacks to be transferrable, it is sufficient if the proxy can approximate the target model in the neighborhood of the harmful query. Therefore, in this paper, we propose \emph{Local Fine-Tuning (LoFT)}, \textit{i.e.}, fine-tuning proxy models on similar queries that lie in the lexico-semantic neighborhood of harmful queries to decrease the divergence between the proxy and target models. First, we demonstrate three approaches to prompt private target models to obtain similar queries given harmful queries. Next, we obtain data for local fine-tuning by eliciting responses from target models for the generated similar queries. Then, we optimize attack suffixes to generate attack prompts and evaluate the impact of our local fine-tuning on the attack's success rate. Experiments show that local fine-tuning of proxy models improves attack transferability and increases attack success rate by $39\%$, $7\%$, and $0.5\%$ (absolute) on target models ChatGPT, GPT-4, and Claude respectively.

Yandong Wen, Weiyang Liu, Yao Feng et al.

In this paper, we focus on a general yet important learning problem, pairwise similarity learning (PSL). PSL subsumes a wide range of important applications, such as open-set face recognition, speaker verification, image retrieval and person re-identification. The goal of PSL is to learn a pairwise similarity function assigning a higher similarity score to positive pairs (i.e., a pair of samples with the same label) than to negative pairs (i.e., a pair of samples with different label). We start by identifying a key desideratum for PSL, and then discuss how existing methods can achieve this desideratum. We then propose a surprisingly simple proxy-free method, called SimPLE, which requires neither feature/proxy normalization nor angular margin and yet is able to generalize well in open-set recognition. We apply the proposed method to three challenging PSL tasks: open-set face recognition, image retrieval and speaker verification. Comprehensive experimental results on large-scale benchmarks show that our method performs significantly better than current state-of-the-art methods.

Thanh-Dat Truong, Utsav Prabhu, Bhiksha Raj et al.

Continual Learning in semantic scene segmentation aims to continually learn new unseen classes in dynamic environments while maintaining previously learned knowledge. Prior studies focused on modeling the catastrophic forgetting and background shift challenges in continual learning. However, fairness, another major challenge that causes unfair predictions leading to low performance among major and minor classes, still needs to be well addressed. In addition, prior methods have yet to model the unknown classes well, thus resulting in producing non-discriminative features among unknown classes. This work presents a novel Fairness Learning via Contrastive Attention Approach to continual learning in semantic scene understanding. In particular, we first introduce a new Fairness Contrastive Clustering loss to address the problems of catastrophic forgetting and fairness. Then, we propose an attention-based visual grammar approach to effectively model the background shift problem and unknown classes, producing better feature representations for different unknown classes. Through our experiments, our proposed approach achieves State-of-the-Art (SoTA) performance on different continual learning benchmarks, i.e., ADE20K, Cityscapes, and Pascal VOC. It promotes the fairness of the continual semantic segmentation model.

Dareen Alharthi, Roshan Sharma, Hira Dhamyal et al.

Modern speech synthesis systems have improved significantly, with synthetic speech being indistinguishable from real speech. However, efficient and holistic evaluation of synthetic speech still remains a significant challenge. Human evaluation using Mean Opinion Score (MOS) is ideal, but inefficient due to high costs. Therefore, researchers have developed auxiliary automatic metrics like Word Error Rate (WER) to measure intelligibility. Prior works focus on evaluating synthetic speech based on pre-trained speech recognition models, however, this can be limiting since this approach primarily measures speech intelligibility. In this paper, we propose an evaluation technique involving the training of an ASR model on synthetic speech and assessing its performance on real speech. Our main assumption is that by training the ASR model on the synthetic speech, the WER on real speech reflects the similarity between distributions, a broader assessment of synthetic speech quality beyond intelligibility. Our proposed metric demonstrates a strong correlation with both MOS naturalness and MOS intelligibility when compared to SpeechLMScore and MOSNet on three recent Text-to-Speech (TTS) systems: MQTTS, StyleTTS, and YourTTS.

Hira Dhamyal, Benjamin Elizalde, Soham Deshmukh et al.

Emotions lie on a continuum, but current models treat emotions as a finite valued discrete variable. This representation does not capture the diversity in the expression of emotion. To better represent emotions we propose the use of natural language descriptions (or prompts). In this work, we address the challenge of automatically generating these prompts and training a model to better learn emotion representations from audio and prompt pairs. We use acoustic properties that are correlated to emotion like pitch, intensity, speech rate, and articulation rate to automatically generate prompts i.e. 'acoustic prompts'. We use a contrastive learning objective to map speech to their respective acoustic prompts. We evaluate our model on Emotion Audio Retrieval and Speech Emotion Recognition. Our results show that the acoustic prompts significantly improve the model's performance in EAR, in various Precision@K metrics. In SER, we observe a 3.8% relative accuracy improvement on the Ravdess dataset.

Yuxuan Wu, Ziyu Wang, Bhiksha Raj et al.

We contribute an unsupervised method that effectively learns disentangled content and style representations from sequences of observations. Unlike most disentanglement algorithms that rely on domain-specific labels or knowledge, our method is based on the insight of domain-general statistical differences between content and style -- content varies more among different fragments within a sample but maintains an invariant vocabulary across data samples, whereas style remains relatively invariant within a sample but exhibits more significant variation across different samples. We integrate such inductive bias into an encoder-decoder architecture and name our method after V3 (variance-versus-invariance). Experimental results show that V3 generalizes across multiple domains and modalities, successfully learning disentangled content and style representations, such as pitch and timbre from music audio, digit and color from images of hand-written digits, and action and character appearance from simple animations. V3 demonstrates strong disentanglement performance compared to existing unsupervised methods, along with superior out-of-distribution generalization under few-shot adaptation compared to supervised counterparts. Lastly, symbolic-level interpretability emerges in the learned content codebook, forging a near one-to-one alignment between machine representation and human knowledge.

Xiang Li, Yinpeng Chen, Chung-Ching Lin et al.

This paper presents a novel approach to object completion, with the primary goal of reconstructing a complete object from its partially visible components. Our method, named MaskComp, delineates the completion process through iterative stages of generation and segmentation. In each iteration, the object mask is provided as an additional condition to boost image generation, and, in return, the generated images can lead to a more accurate mask by fusing the segmentation of images. We demonstrate that the combination of one generation and one segmentation stage effectively functions as a mask denoiser. Through alternation between the generation and segmentation stages, the partial object mask is progressively refined, providing precise shape guidance and yielding superior object completion results. Our experiments demonstrate the superiority of MaskComp over existing approaches, e.g., ControlNet and Stable Diffusion, establishing it as an effective solution for object completion.