Audio embeddings are crucial tools in understanding large catalogs of music. Typically embeddings are evaluated on the basis of the performance they provide in a wide range of downstream tasks, however few studies have investigated the local properties of the embedding spaces themselves which are important in nearest neighbor algorithms, commonly used in music search and recommendation. In this work we show that when learning audio representations on music datasets via contrastive learning, musical properties that are typically homogeneous within a track (e.g., key and tempo) are reflected in the locality of neighborhoods in the resulting embedding space. By applying appropriate data augmentation strategies, localisation of such properties can not only be reduced but the localisation of other attributes is increased. For example, locality of features such as pitch and tempo that are less relevant to non-expert listeners, may be mitigated while improving the locality of more salient features such as genre and mood, achieving state-of-the-art performance in nearest neighbor retrieval accuracy. Similarly, we show that the optimal selection of data augmentation strategies for contrastive learning of music audio embeddings is dependent on the downstream task, highlighting this as an important embedding design decision.
This paper introduces a novel approach named CrossVideo, which aims to enhance self-supervised cross-modal contrastive learning in the field of point cloud video understanding. Traditional supervised learning methods encounter limitations due to data scarcity and challenges in label acquisition. To address these issues, we propose a self-supervised learning method that leverages the cross-modal relationship between point cloud videos and image videos to acquire meaningful feature representations. Intra-modal and cross-modal contrastive learning techniques are employed to facilitate effective comprehension of point cloud video. We also propose a multi-level contrastive approach for both modalities. Through extensive experiments, we demonstrate that our method significantly surpasses previous state-of-the-art approaches, and we conduct comprehensive ablation studies to validate the effectiveness of our proposed designs.
The generation of undesirable and factually incorrect content of large language models poses a significant challenge and remains largely an unsolved issue. This paper studies the integration of a contrastive learning objective for fine-tuning LLMs for implicit knowledge editing and controlled text generation. Optimizing the training objective entails aligning text perplexities in a contrastive fashion. To facilitate training the model in a self-supervised fashion, we leverage an off-the-shelf LLM for training data generation. We showcase applicability in the domain of detoxification. Herein, the proposed approach leads to a significant decrease in the generation of toxic content while preserving general utility for downstream tasks such as commonsense reasoning and reading comprehension. The proposed approach is conceptually simple but empirically powerful.
Explaining multivariate time series is a compound challenge, as it requires identifying important locations in the time series and matching complex temporal patterns. Although previous saliency-based methods addressed the challenges, their perturbation may not alleviate the distribution shift issue, which is inevitable especially in heterogeneous samples. We present ContraLSP, a locally sparse model that introduces counterfactual samples to build uninformative perturbations but keeps distribution using contrastive learning. Furthermore, we incorporate sample-specific sparse gates to generate more binary-skewed and smooth masks, which easily integrate temporal trends and select the salient features parsimoniously. Empirical studies on both synthetic and real-world datasets show that ContraLSP outperforms state-of-the-art models, demonstrating a substantial improvement in explanation quality for time series data. The source code is available at \url{https://github.com/zichuan-liu/ContraLSP}.
Existing unsupervised deep product quantization methods primarily aim for the increased similarity between different views of the identical image, whereas the delicate multi-level semantic similarities preserved between images are overlooked. Moreover, these methods predominantly focus on the Euclidean space for computational convenience, compromising their ability to map the multi-level semantic relationships between images effectively. To mitigate these shortcomings, we propose a novel unsupervised product quantization method dubbed \textbf{Hi}erarchical \textbf{H}yperbolic \textbf{P}roduct \textbf{Q}uantization (HiHPQ), which learns quantized representations by incorporating hierarchical semantic similarity within hyperbolic geometry. Specifically, we propose a hyperbolic product quantizer, where the hyperbolic codebook attention mechanism and the quantized contrastive learning on the hyperbolic product manifold are introduced to expedite quantization. Furthermore, we propose a hierarchical semantics learning module, designed to enhance the distinction between similar and non-matching images for a query by utilizing the extracted hierarchical semantics as an additional training supervision. Experiments on benchmarks show that our proposed method outperforms state-of-the-art baselines.
Customizable keyword spotting (KWS) in continuous speech has attracted increasing attention due to its real-world application potential. While contrastive learning (CL) has been widely used to extract keyword representations, previous CL approaches all operate on pre-segmented isolated words and employ only audio-text representations matching strategy. However, for KWS in continuous speech, co-articulation and streaming word segmentation can easily yield similar audio patterns for different texts, which may consequently trigger false alarms. To address this issue, we propose a novel CL with Audio Discrimination (CLAD) approach to learning keyword representation with both audio-text matching and audio-audio discrimination ability. Here, an InfoNCE loss considering both audio-audio and audio-text CL data pairs is employed for each sliding window during training. Evaluations on the open-source LibriPhrase dataset show that the use of sliding-window level InfoNCE loss yields comparable performance compared to previous CL approaches. Furthermore, experiments on the continuous speech dataset LibriSpeech demonstrate that, by incorporating audio discrimination, CLAD achieves significant performance gain over CL without audio discrimination. Meanwhile, compared to two-stage KWS approaches, the end-to-end KWS with CLAD achieves not only better performance, but also significant speed-up.
Monitoring organic matter is pivotal for maintaining soil health and can help inform sustainable soil management practices. While sensor-based soil information offers higher-fidelity and reliable insights into organic matter changes, sampling and measuring sensor data is cost-prohibitive. We propose a multi-modal, scalable framework that can estimate organic matter from remote sensing data, a more readily available data source while leveraging sparse soil information for improving generalization. Using the sensor data, we preserve underlying causal relations among sensor attributes and organic matter. Simultaneously we leverage inherent structure in the data and train the model to discriminate among domains using contrastive learning. This causal and contrastive constraint minimization ensures improved generalization and adaptation to other domains. We also shed light on the interpretability of the framework by identifying attributes that are important for improving generalization. Identifying these key soil attributes that affect organic matter will aid in efforts to standardize data collection efforts.
Cutting-edge research in facial expression recognition (FER) currently favors the utilization of convolutional neural networks (CNNs) backbone which is supervisedly pre-trained on face recognition datasets for feature extraction. However, due to the vast scale of face recognition datasets and the high cost associated with collecting facial labels, this pre-training paradigm incurs significant expenses. Towards this end, we propose to pre-train vision Transformers (ViTs) through a self-supervised approach on a mid-scale general image dataset. In addition, when compared with the domain disparity existing between face datasets and FER datasets, the divergence between general datasets and FER datasets is more pronounced. Therefore, we propose a contrastive fine-tuning approach to effectively mitigate this domain disparity. Specifically, we introduce a novel FER training paradigm named Mask Image pre-training with MIx Contrastive fine-tuning (MIMIC). In the initial phase, we pre-train the ViT via masked image reconstruction on general images. Subsequently, in the fine-tuning stage, we introduce a mix-supervised contrastive learning process, which enhances the model with a more extensive range of positive samples by the mixing strategy. Through extensive experiments conducted on three benchmark datasets, we demonstrate that our MIMIC outperforms the previous training paradigm, showing its capability to learn better representations. Remarkably, the results indicate that the vanilla ViT can achieve impressive performance without the need for intricate, auxiliary-designed modules. Moreover, when scaling up the model size, MIMIC exhibits no performance saturation and is superior to the current state-of-the-art methods.
Vein recognition has received increasing attention due to its high security and privacy. Recently, deep neural networks such as Convolutional neural networks (CNN) and Transformers have been introduced for vein recognition and achieved state-of-the-art performance. Despite the recent advances, however, existing solutions for finger-vein feature extraction are still not optimal due to scarce training image samples. To overcome this problem, in this paper, we propose an adversarial masking contrastive learning (AMCL) approach, that generates challenging samples to train a more robust contrastive learning model for the downstream palm-vein recognition task, by alternatively optimizing the encoder in the contrastive learning model and a set of latent variables. First, a huge number of masks are generated to train a robust generative adversarial network (GAN). The trained generator transforms a latent variable from the latent variable space into a mask space. Then, we combine the trained generator with a contrastive learning model to obtain our AMCL, where the generator produces challenging masking images to increase the contrastive loss and the contrastive learning model is trained based on the harder images to learn a more robust feature representation. After training, the trained encoder in the contrastive learning model is combined with a classification layer to build a classifier, which is further fine-tuned on labeled training data for vein recognition. The experimental results on three databases demonstrate that our approach outperforms existing contrastive learning approaches in terms of improving identification accuracy of vein classifiers and achieves state-of-the-art recognition results.
Noisy labels can significantly impact medical image classification, particularly in deep learning, by corrupting learned features. Self-supervised pretraining, which doesn't rely on labeled data, can enhance robustness against noisy labels. However, this robustness varies based on factors like the number of classes, dataset complexity, and training size. In medical images, subtle inter-class differences and modality-specific characteristics add complexity. Previous research hasn't comprehensively explored the interplay between self-supervised learning and robustness against noisy labels in medical image classification, considering all these factors. In this study, we address three key questions: i) How does label noise impact various medical image classification datasets? ii) Which types of medical image datasets are more challenging to learn and more affected by label noise? iii) How do different self-supervised pretraining methods enhance robustness across various medical image datasets? Our results show that DermNet, among five datasets (Fetal plane, DermNet, COVID-DU-Ex, MURA, NCT-CRC-HE-100K), is the most challenging but exhibits greater robustness against noisy labels. Additionally, contrastive learning stands out among the eight self-supervised methods as the most effective approach to enhance robustness against noisy labels.
Vertebrae identification in arbitrary fields-of-view plays a crucial role in diagnosing spine disease. Most spine CT contain only local regions, such as the neck, chest, and abdomen. Therefore, identification should not depend on specific vertebrae or a particular number of vertebrae being visible. Existing methods at the spine-level are unable to meet this challenge. In this paper, we propose a three-stage method to address the challenges in 3D CT vertebrae identification at vertebrae-level. By sequentially performing the tasks of vertebrae localization, segmentation, and identification, the anatomical prior information of the vertebrae is effectively utilized throughout the process. Specifically, we introduce a dual-factor density clustering algorithm to acquire localization information for individual vertebra, thereby facilitating subsequent segmentation and identification processes. In addition, to tackle the issue of interclass similarity and intra-class variability, we pre-train our identification network by using a supervised contrastive learning method. To further optimize the identification results, we estimated the uncertainty of the classification network and utilized the message fusion module to combine the uncertainty scores, while aggregating global information about the spine. Our method achieves state-of-the-art results on the VerSe19 and VerSe20 challenge benchmarks. Additionally, our approach demonstrates outstanding generalization performance on an collected dataset containing a wide range of abnormal cases.
The world's more than 7000 languages are written in at least 293 scripts. Due to various reasons, many closely related languages use different scripts, which poses a difficulty for multilingual pretrained language models (mPLMs) in learning crosslingual knowledge through lexical overlap. As a consequence, mPLMs are faced with a script barrier: representations from different scripts are located in different subspaces, which can result in crosslingual transfer involving languages of different scripts performing suboptimally. To address this problem, we propose TransliCo, a framework that optimizes the Transliteration Contrastive Modeling (TCM) objective to fine-tune an mPLM by contrasting sentences in its training data and their transliterations in a unified script (in our case Latin), which enhances uniformity in the representation space for different scripts. Using Glot500-m, an mPLM pretrained on over 500 languages, as our source model, we fine-tune it on a small portion (5%) of its training data, and refer to the resulting model as Furina. We show that Furina not only better aligns representations from distinct scripts but also outperforms the original Glot500-m on various zero-shot crosslingual transfer tasks. Additionally, we achieve consistent improvement in a case study on the Indic group where the languages exhibit areal features but use different scripts. We make our code and models publicly available.
Federated clustering, an essential extension of centralized clustering for federated scenarios, enables multiple data-holding clients to collaboratively group data while keeping their data locally. In centralized scenarios, clustering driven by representation learning has made significant advancements in handling high-dimensional complex data. However, the combination of federated clustering and representation learning remains underexplored. To bridge this, we first tailor a cluster-contrastive model for learning clustering-friendly representations. Then, we harness this model as the foundation for proposing a new federated clustering method, named cluster-contrastive federated clustering (CCFC). Benefiting from representation learning, the clustering performance of CCFC even double those of the best baseline methods in some cases. Compared to the most related baseline, the benefit results in substantial NMI score improvements of up to 0.4155 on the most conspicuous case. Moreover, CCFC also shows superior performance in handling device failures from a practical viewpoint.
An emotional support conversation system aims to alleviate users' emotional distress and assist them in addressing their challenges. To generate supportive responses, it is critical to consider multiple factors such as empathy, support strategies, and response coherence, as established in prior methods. Nonetheless, previous models occasionally generate unhelpful responses, which intend to provide support but display counterproductive effects. According to psychology and communication theories, poor performance in just one contributing factor might cause a response to be unhelpful. From the model training perspective, since these models have not been exposed to unhelpful responses during their training phase, they are unable to distinguish if the tokens they generate might result in unhelpful responses during inference. To address this issue, we introduce a novel model-agnostic framework named mitigating unhelpfulness with multifaceted AI feedback for emotional support (Muffin). Specifically, Muffin employs a multifaceted AI feedback module to assess the helpfulness of responses generated by a specific model with consideration of multiple factors. Using contrastive learning, it then reduces the likelihood of the model generating unhelpful responses compared to the helpful ones. Experimental results demonstrate that Muffin effectively mitigates the generation of unhelpful responses while slightly increasing response fluency and relevance.
Leveraging wearable devices for motion reconstruction has emerged as an economical and viable technique. Certain methodologies employ sparse Inertial Measurement Units (IMUs) on the human body and harness data-driven strategies to model human poses. However, the reconstruction of motion based solely on sparse IMUs data is inherently fraught with ambiguity, a consequence of numerous identical IMU readings corresponding to different poses. In this paper, we explore the spatial importance of multiple sensors, supervised by text that describes specific actions. Specifically, uncertainty is introduced to derive weighted features for each IMU. We also design a Hierarchical Temporal Transformer (HTT) and apply contrastive learning to achieve precise temporal and feature alignment of sensor data with textual semantics. Experimental results demonstrate our proposed approach achieves significant improvements in multiple metrics compared to existing methods. Notably, with textual supervision, our method not only differentiates between ambiguous actions such as sitting and standing but also produces more precise and natural motion.
Retinal image registration is of utmost importance due to its wide applications in medical practice. In this context, we propose ConKeD, a novel deep learning approach to learn descriptors for retinal image registration. In contrast to current registration methods, our approach employs a novel multi-positive multi-negative contrastive learning strategy that enables the utilization of additional information from the available training samples. This makes it possible to learn high quality descriptors from limited training data. To train and evaluate ConKeD, we combine these descriptors with domain-specific keypoints, particularly blood vessel bifurcations and crossovers, that are detected using a deep neural network. Our experimental results demonstrate the benefits of the novel multi-positive multi-negative strategy, as it outperforms the widely used triplet loss technique (single-positive and single-negative) as well as the single-positive multi-negative alternative. Additionally, the combination of ConKeD with the domain-specific keypoints produces comparable results to the state-of-the-art methods for retinal image registration, while offering important advantages such as avoiding pre-processing, utilizing fewer training samples, and requiring fewer detected keypoints, among others. Therefore, ConKeD shows a promising potential towards facilitating the development and application of deep learning-based methods for retinal image registration.
Accurate lesion classification in Wireless Capsule Endoscopy (WCE) images is vital for early diagnosis and treatment of gastrointestinal (GI) cancers. However, this task is confronted with challenges like tiny lesions and background interference. Additionally, WCE images exhibit higher intra-class variance and inter-class similarities, adding complexity. To tackle these challenges, we propose Decoupled Supervised Contrastive Learning for WCE image classification, learning robust representations from zoomed-in WCE images generated by Saliency Augmentor. Specifically, We use uniformly down-sampled WCE images as anchors and WCE images from the same class, especially their zoomed-in images, as positives. This approach empowers the Feature Extractor to capture rich representations from various views of the same image, facilitated by Decoupled Supervised Contrastive Learning. Training a linear Classifier on these representations within 10 epochs yields an impressive 92.01% overall accuracy, surpassing the prior state-of-the-art (SOTA) by 0.72% on a blend of two publicly accessible WCE datasets. Code is available at: https://github.com/Qiukunpeng/DSCL.
Audio-Visual Emotion Recognition (AVER) has garnered increasing attention in recent years for its critical role in creating emotion-ware intelligent machines. Previous efforts in this area are dominated by the supervised learning paradigm. Despite significant progress, supervised learning is meeting its bottleneck due to the longstanding data scarcity issue in AVER. Motivated by recent advances in self-supervised learning, we propose Hierarchical Contrastive Masked Autoencoder (HiCMAE), a novel self-supervised framework that leverages large-scale self-supervised pre-training on vast unlabeled audio-visual data to promote the advancement of AVER. Following prior arts in self-supervised audio-visual representation learning, HiCMAE adopts two primary forms of self-supervision for pre-training, namely masked data modeling and contrastive learning. Unlike them which focus exclusively on top-layer representations while neglecting explicit guidance of intermediate layers, HiCMAE develops a three-pronged strategy to foster hierarchical audio-visual feature learning and improve the overall quality of learned representations. To verify the effectiveness of HiCMAE, we conduct extensive experiments on 9 datasets covering both categorical and dimensional AVER tasks. Experimental results show that our method significantly outperforms state-of-the-art supervised and self-supervised audio-visual methods, which indicates that HiCMAE is a powerful audio-visual emotion representation learner. Codes and models will be publicly available at https://github.com/sunlicai/HiCMAE.
Few-shot Named Entity Recognition (NER) task, which aims to identify and classify entities from different domains with limited training samples, has long been treated as a basic step for knowledge graph (KG) construction. Great efforts have been made on this task with competitive performance, however, they usually treat the two subtasks, namely span detection and type classification, as mutually independent, and the integrity and correlation between subtasks have been largely ignored. Moreover, prior arts may fail to absorb the coarse-grained features of entities, resulting in a semantic-insufficient representation of entity types. To that end, in this paper, we propose a Hierarchical Enhancing ProtoNet (HEProto) based on multi-task learning, which is utilized to jointly learn these two subtasks and model their correlation. Specifically, we adopt contrastive learning to enhance the span boundary information and the type semantic representations in these two subtasks. Then, the hierarchical prototypical network is designed to leverage the coarse-grained information of entities in the type classification stage, which could help the model to better learn the fine-grained semantic representations. Along this line, we construct a similarity margin loss to reduce the similarity between fine-grained entities and other irrelevant coarse-grained prototypes. Finally, extensive experiments on the Few-NERD dataset prove that our solution outperforms competitive baseline methods. The source code of HEProto is available at https://github.com/fanshu6hao/HEProto.
Reinforcement Learning from Human Feedback (RLHF) has become a crucial technology for aligning language models with human values and intentions, enabling models to produce more helpful and harmless responses. Reward models are trained as proxies for human preferences to drive reinforcement learning optimization. While reward models are often considered central to achieving high performance, they face the following challenges in practical applications: (1) Incorrect and ambiguous preference pairs in the dataset may hinder the reward model from accurately capturing human intent. (2) Reward models trained on data from a specific distribution often struggle to generalize to examples outside that distribution and are not suitable for iterative RLHF training. In this report, we attempt to address these two issues. (1) From a data perspective, we propose a method to measure the strength of preferences within the data, based on a voting mechanism of multiple reward models. Experimental results confirm that data with varying preference strengths have different impacts on reward model performance. We introduce a series of novel methods to mitigate the influence of incorrect and ambiguous preferences in the dataset and fully leverage high-quality preference data. (2) From an algorithmic standpoint, we introduce contrastive learning to enhance the ability of reward models to distinguish between chosen and rejected responses, thereby improving model generalization. Furthermore, we employ meta-learning to enable the reward model to maintain the ability to differentiate subtle differences in out-of-distribution samples, and this approach can be utilized for iterative RLHF optimization.