Multi-modality image fusion is a technique that combines information from different sensors or modalities, enabling the fused image to retain complementary features from each modality, such as functional highlights and texture details. However, effective training of such fusion models is challenging due to the scarcity of ground truth fusion data. To tackle this issue, we propose the Equivariant Multi-Modality imAge fusion (EMMA) paradigm for end-to-end self-supervised learning. Our approach is rooted in the prior knowledge that natural imaging responses are equivariant to certain transformations. Consequently, we introduce a novel training paradigm that encompasses a fusion module, a pseudo-sensing module, and an equivariant fusion module. These components enable the net training to follow the principles of the natural sensing-imaging process while satisfying the equivariant imaging prior. Extensive experiments confirm that EMMA yields high-quality fusion results for infrared-visible and medical images, concurrently facilitating downstream multi-modal segmentation and detection tasks. The code is available at https://github.com/Zhaozixiang1228/MMIF-EMMA.
Existing solutions to zero-shot text classification either conduct prompting with pre-trained language models, which is sensitive to the choices of templates, or rely on large-scale annotated data of relevant tasks for meta-tuning. In this work, we propose a new paradigm based on self-supervised learning to solve zero-shot text classification tasks by tuning the language models with unlabeled data, called self-supervised tuning. By exploring the inherent structure of free texts, we propose a new learning objective called first sentence prediction to bridge the gap between unlabeled data and text classification tasks. After tuning the model to learn to predict the first sentence in a paragraph based on the rest, the model is able to conduct zero-shot inference on unseen tasks such as topic classification and sentiment analysis. Experimental results show that our model outperforms the state-of-the-art baselines on 7 out of 10 tasks. Moreover, the analysis reveals that our model is less sensitive to the prompt design. Our code and pre-trained models are publicly available at https://github.com/DAMO-NLP-SG/SSTuning .
Self-supervised speech representation models have succeeded in various tasks, but improving them for content-related problems using unlabeled data is challenging. We propose speaker-invariant clustering (Spin), a novel self-supervised learning method that clusters speech representations and performs swapped prediction between the original and speaker-perturbed utterances. Spin disentangles speaker information and preserves content representations with just 45 minutes of fine-tuning on a single GPU. Spin improves pre-trained networks and outperforms prior methods in speech recognition and acoustic unit discovery.
Graph neural networks (GNNs) have recently emerged as an effective collaborative filtering (CF) approaches for recommender systems. The key idea of GNN-based recommender systems is to recursively perform message passing along user-item interaction edges to refine encoded embeddings, relying on sufficient and high-quality training data. However, user behavior data in practical recommendation scenarios is often noisy and exhibits skewed distribution. To address these issues, some recommendation approaches, such as SGL, leverage self-supervised learning to improve user representations. These approaches conduct self-supervised learning through creating contrastive views, but they depend on the tedious trial-and-error selection of augmentation methods. In this paper, we propose a novel Adaptive Graph Contrastive Learning (AdaGCL) framework that conducts data augmentation with two adaptive contrastive view generators to better empower the CF paradigm. Specifically, we use two trainable view generators - a graph generative model and a graph denoising model - to create adaptive contrastive views. With two adaptive contrastive views, AdaGCL introduces additional high-quality training signals into the CF paradigm, helping to alleviate data sparsity and noise issues. Extensive experiments on three real-world datasets demonstrate the superiority of our model over various state-of-the-art recommendation methods. Our model implementation codes are available at the link https://github.com/HKUDS/AdaGCL.
The current monaural state of the art tools for speech separation relies on supervised learning. This means that they must deal with permutation problem, they are impacted by the mismatch on the number of speakers used in training and inference. Moreover, their performance heavily relies on the presence of high-quality labelled data. These problems can be effectively addressed by employing a fully unsupervised technique for speech separation. In this paper, we use contrastive learning to establish the representations of frames then use the learned representations in the downstream deep modularization task. Concretely, we demonstrate experimentally that in speech separation, different frames of a speaker can be viewed as augmentations of a given hidden standard frame of that speaker. The frames of a speaker contain enough prosodic information overlap which is key in speech separation. Based on this, we implement a self-supervised learning to learn to minimize the distance between frames belonging to a given speaker. The learned representations are used in a downstream deep modularization task to cluster frames based on speaker identity. Evaluation of the developed technique on WSJ0-2mix and WSJ0-3mix shows that the technique attains SI-SNRi and SDRi of 20.8 and 21.0 respectively in WSJ0-2mix. In WSJ0-3mix, it attains SI-SNRi and SDRi of 20.7 and 20.7 respectively in WSJ0-2mix. Its greatest strength being that as the number of speakers increase, its performance does not degrade significantly.
Speech processing Universal PERformance Benchmark (SUPERB) is a leaderboard to benchmark the performance of Self-Supervised Learning (SSL) models on various speech processing tasks. However, SUPERB largely considers English speech in its evaluation. This paper presents multilingual SUPERB (ML-SUPERB), covering 143 languages (ranging from high-resource to endangered), and considering both automatic speech recognition and language identification. Following the concept of SUPERB, ML-SUPERB utilizes frozen SSL features and employs a simple framework for multilingual tasks by learning a shallow downstream model. Similar to the SUPERB benchmark, we find speech SSL models can significantly improve performance compared to FBANK features. Furthermore, we find that multilingual models do not always perform better than their monolingual counterparts. We will release ML-SUPERB as a challenge with organized datasets and reproducible training scripts for future multilingual representation research.
Dynamic positron emission tomography imaging (dPET) provides temporally resolved images of a tracer enabling a quantitative measure of physiological processes. Voxel-wise physiologically-based pharmacokinetic (PBPK) modeling of the time activity curves (TAC) can provide relevant diagnostic information for clinical workflow. Conventional fitting strategies for TACs are slow and ignore the spatial relation between neighboring voxels. We train a spatio-temporal UNet to estimate the kinetic parameters given TAC from F-18-fluorodeoxyglucose (FDG) dPET. This work introduces a self-supervised loss formulation to enforce the similarity between the measured TAC and those generated with the learned kinetic parameters. Our method provides quantitatively comparable results at organ-level to the significantly slower conventional approaches, while generating pixel-wise parametric images which are consistent with expected physiology. To the best of our knowledge, this is the first self-supervised network that allows voxel-wise computation of kinetic parameters consistent with a non-linear kinetic model. The code will become publicly available upon acceptance.
Animal-borne sensors (`bio-loggers') can record a suite of kinematic and environmental data, which are used to elucidate animal ecophysiology and improve conservation efforts. Machine learning techniques are used for interpreting the large amounts of data recorded by bio-loggers, but there exists no common framework for comparing the different machine learning techniques in this domain. This makes it difficult to, for example, identify patterns in what works well for machine learning-based analysis of bio-logger data. It also makes it difficult to evaluate the effectiveness of novel methods developed by the machine learning community. To address this, we present the Bio-logger Ethogram Benchmark (BEBE), a collection of datasets with behavioral annotations, as well as a modeling task and evaluation metrics. BEBE is to date the largest, most taxonomically diverse, publicly available benchmark of this type. Using BEBE, we compare the performance of deep and classical machine learning methods for identifying animal behaviors based on bio-logger data. As an example usage of BEBE, we test an approach based on self-supervised learning. To apply this approach to animal behavior classification, we adapt a deep neural network pre-trained with 700,000 hours of data collected from human wrist-worn accelerometers. We find that deep neural networks out-perform the classical machine learning methods we tested across all nine datasets in BEBE. We additionally find that the approach based on self-supervised learning out-performs the alternatives we tested, especially in settings when there is a low amount of training data available. In light of this, we are able to make concrete suggestions for designing studies that rely on machine learning to infer behavior from bio-logger data. Datasets and code are available at https://github.com/earthspecies/BEBE.
In recent times, contrastive learning based loss functions have become increasingly popular for visual self-supervised representation learning owing to their state-of-the-art (SOTA) performance. Most of the modern contrastive learning methods generalize only to one positive and multiple negatives per anchor. A recent state-of-the-art, supervised contrastive (SupCon) loss, extends self-supervised contrastive learning to supervised setting by generalizing to multiple positives and negatives in a batch and improves upon the cross-entropy loss. In this paper, we propose a novel contrastive loss function -- Tuned Contrastive Learning (TCL) loss, that generalizes to multiple positives and negatives in a batch and offers parameters to tune and improve the gradient responses from hard positives and hard negatives. We provide theoretical analysis of our loss function's gradient response and show mathematically how it is better than that of SupCon loss. We empirically compare our loss function with SupCon loss and cross-entropy loss in supervised setting on multiple classification-task datasets to show its effectiveness. We also show the stability of our loss function to a range of hyper-parameter settings. Unlike SupCon loss which is only applied to supervised setting, we show how to extend TCL to self-supervised setting and empirically compare it with various SOTA self-supervised learning methods. Hence, we show that TCL loss achieves performance on par with SOTA methods in both supervised and self-supervised settings.
Hyperbolic manifolds for visual representation learning allow for effective learning of semantic class hierarchies by naturally embedding tree-like structures with low distortion within a low-dimensional representation space. The highly separable semantic class hierarchies produced by hyperbolic learning have shown to be powerful in low-shot tasks, however, their application in self-supervised learning is yet to be explored fully. In this work, we explore the use of hyperbolic representation space for self-supervised representation learning for prototype-based clustering approaches. First, we extend the Masked Siamese Networks to operate on the Poincar\'e ball model of hyperbolic space, secondly, we place prototypes on the ideal boundary of the Poincar\'e ball. Unlike previous methods we project to the hyperbolic space at the output of the encoder network and utilise a hyperbolic projection head to ensure that the representations used for downstream tasks remain hyperbolic. Empirically we demonstrate the ability of these methods to perform comparatively to Euclidean methods in lower dimensions for linear evaluation tasks, whilst showing improvements in extreme few-shot learning tasks.
Graph anomaly detection has long been an important problem in various domains pertaining to information security such as financial fraud, social spam and network intrusion. The majority of existing methods are performed in an unsupervised manner, as labeled anomalies in a large scale are often too expensive to acquire. However, the identified anomalies may turn out to be uninteresting data instances due to the lack of prior knowledge. In real-world scenarios, it is often feasible to obtain limited labeled anomalies, which have great potential to advance graph anomaly detection. However, the work exploring limited labeled anomalies and a large amount of unlabeled nodes in graphs to detect anomalies is relatively limited. Therefore, in this paper, we study an important problem of few-shot graph anomaly detection. Nonetheless, it is challenging to fully leverage the information of few-shot anomalous nodes due to the irregularity of anomalies and the overfitting issue in the few-shot learning. To tackle the above challenges, we propose a novel meta-learning based framework, MetaGAD, that learns to adapt the knowledge from self-supervised learning to few-shot supervised learning for graph anomaly detection. In specific, we formulate the problem as a bi-level optimization, ensuring MetaGAD converging to minimizing the validation loss, thus enhancing the generalization capacity. The comprehensive experiments on six real-world datasets with synthetic anomalies and "organic" anomalies (available in the datasets) demonstrate the effectiveness of MetaGAD in detecting anomalies with few-shot anomalies. The code is available at https://github.com/XiongxiaoXu/MetaGAD.
We consider a contrastive learning approach to knowledge graph embedding (KGE) via InfoNCE. For KGE, efficient learning relies on augmenting the training data with negative triples. However, most KGE works overlook the bias from generating the negative triples-false negative triples (factual triples missing from the knowledge graph). We argue that the generation of high-quality (i.e., hard) negative triples might lead to an increase in false negative triples. To mitigate the impact of false negative triples during the generation of hard negative triples, we propose the Hardness and Structure-aware (\textbf{HaSa}) contrastive KGE method, which alleviates the effect of false negative triples while generating the hard negative triples. Experiments show that HaSa improves the performance of InfoNCE-based KGE approaches and achieves state-of-the-art results in several metrics for WN18RR datasets and competitive results for FB15k-237 datasets compared to both classic and pre-trained LM-based KGE methods.
Extracting information from the electrocardiography (ECG) signal is an essential step in the design of digital health technologies in cardiology. In recent years, several machine learning (ML) algorithms for automatic extraction of information in ECG have been proposed. Supervised learning methods have successfully been used to identify specific aspects in the signal, like detection of rhythm disorders (arrhythmias). Self-supervised learning (SSL) methods, on the other hand, can be used to extract all the features contained in the data. The model is optimized without any specific goal and learns from the data itself. By adapting state-of-the-art computer vision methodologies to the signal processing domain, a few SSL approaches have been reported recently for ECG processing. However, such SSL methods require either data augmentation or negative pairs, which limits the method to only look for similarities between two ECG inputs, either two versions of the same signal or two signals from the same subject. This leads to models that are very effective at extracting characteristics that are stable in a subject, such as gender or age. But they are not successful at capturing changes within the ECG recording that can explain dynamic aspects, like different arrhythmias or different sleep stages. In this work, we introduce the first SSL method that uses neither data augmentation nor negative pairs for understanding ECG signals, and still, achieves comparable quality representations. As a result, it is possible to design a SSL method that not only captures similarities between two inputs, but also captures dissimilarities for a complete understanding of the data. In addition, a model based on transformer blocks is presented, which produces better results than a model based on convolutional layers (XResNet50) with almost the same number of parameters.
The manifold hypothesis posits that high-dimensional data often lies on a lower-dimensional manifold and that utilizing this manifold as the target space yields more efficient representations. While numerous traditional manifold-based techniques exist for dimensionality reduction, their application in self-supervised learning has witnessed slow progress. The recent MSimCLR method combines manifold encoding with SimCLR but requires extremely low target encoding dimensions to outperform SimCLR, limiting its applicability. This paper introduces a novel learning paradigm using an unbalanced atlas (UA), capable of surpassing state-of-the-art self-supervised learning approaches. We investigated and engineered the DeepInfomax with an unbalanced atlas (DIM-UA) method by adapting the Spatiotemporal DeepInfomax (ST-DIM) framework to align with our proposed UA paradigm. The efficacy of DIM-UA is demonstrated through training and evaluation on the Atari Annotated RAM Interface (AtariARI) benchmark, a modified version of the Atari 2600 framework that produces annotated image samples for representation learning. The UA paradigm improves existing algorithms significantly as the number of target encoding dimensions grows. For instance, the mean F1 score averaged over categories of DIM-UA is ~75% compared to ~70% of ST-DIM when using 16384 hidden units.
The so-called black-box deep learning (DL) models are increasingly used in classification tasks across many scientific disciplines, including wireless communications domain. In this trend, supervised DL models appear as most commonly proposed solutions to domain-related classification problems. Although they are proven to have unmatched performance, the necessity for large labeled training data and their intractable reasoning, as two major drawbacks, are constraining their usage. The self-supervised architectures emerged as a promising solution that reduces the size of the needed labeled data, but the explainability problem remains. In this paper, we propose a methodology for explaining deep clustering, self-supervised learning architectures comprised of a representation learning part based on a Convolutional Neural Network (CNN) and a clustering part. For the state of the art representation learning part, our methodology employs Guided Backpropagation to interpret the regions of interest of the input data. For the clustering part, the methodology relies on Shallow Trees to explain the clustering result using optimized depth decision tree. Finally, a data-specific visualizations part enables connection for each of the clusters to the input data trough the relevant features. We explain on a use case of wireless spectrum activity clustering how the CNN-based, deep clustering architecture reasons.
Self-supervised learning aims to extract meaningful features from unlabeled data for further downstream tasks. In this paper, we consider classification as a downstream task in phase 2 and develop rigorous theories to realize the factors that implicitly influence the general loss of this classification task. Our theories signify that sharpness-aware feature extractors benefit the classification task in phase 2 and the existing data shift between the ideal (i.e., the ideal one used in theory development) and practical (i.e., the practical one used in implementation) distributions to generate positive pairs also remarkably affects this classification task. Further harvesting these theoretical findings, we propose to minimize the sharpness of the feature extractor and a new Fourier-based data augmentation technique to relieve the data shift in the distributions generating positive pairs, reaching Sharpness & Shift-Aware Contrastive Learning (SSA-CLR). We conduct extensive experiments to verify our theoretical findings and demonstrate that sharpness & shift-aware contrastive learning can remarkably boost the performance as well as obtaining more robust extracted features compared with the baselines.
In many machine learning applications, labeling datasets can be an arduous and time-consuming task. Although research has shown that semi-supervised learning techniques can achieve high accuracy with very few labels within the field of computer vision, little attention has been given to how images within a dataset should be selected for labeling. In this paper, we propose a novel approach based on well-established self-supervised learning, clustering, and manifold learning techniques that address this challenge of selecting an informative image subset to label in the first instance, which is known as the cold-start or unsupervised selective labelling problem. We test our approach using several publicly available datasets, namely CIFAR10, Imagenette, DeepWeeds, and EuroSAT, and observe improved performance with both supervised and semi-supervised learning strategies when our label selection strategy is used, in comparison to random sampling. We also obtain superior performance for the datasets considered with a much simpler approach compared to other methods in the literature.
Tabular data is the most widely used data format in machine learning (ML). While tree-based methods outperform DL-based methods in supervised learning, recent literature reports that self-supervised learning with Transformer-based models outperforms tree-based methods. In the existing literature on self-supervised learning for tabular data, contrastive learning is the predominant method. In contrastive learning, data augmentation is important to generate different views. However, data augmentation for tabular data has been difficult due to the unique structure and high complexity of tabular data. In addition, three main components are proposed together in existing methods: model structure, self-supervised learning methods, and data augmentation. Therefore, previous works have compared the performance without comprehensively considering these components, and it is not clear how each component affects the actual performance. In this study, we focus on data augmentation to address these issues. We propose a novel data augmentation method, $\textbf{M}$ask $\textbf{T}$oken $\textbf{R}$eplacement ($\texttt{MTR}$), which replaces the mask token with a portion of each tokenized column; $\texttt{MTR}$ takes advantage of the properties of Transformer, which is becoming the predominant DL-based architecture for tabular data, to perform data augmentation for each column embedding. Through experiments with 13 diverse public datasets in both supervised and self-supervised learning scenarios, we show that $\texttt{MTR}$ achieves competitive performance against existing data augmentation methods and improves model performance. In addition, we discuss specific scenarios in which $\texttt{MTR}$ is most effective and identify the scope of its application. The code is available at https://github.com/somaonishi/MTR/.
Nowadays, recognition-synthesis-based methods have been quite popular with voice conversion (VC). By introducing linguistics features with good disentangling characters extracted from an automatic speech recognition (ASR) model, the VC performance achieved considerable breakthroughs. Recently, self-supervised learning (SSL) methods trained with a large-scale unannotated speech corpus have been applied to downstream tasks focusing on the content information, which is suitable for VC tasks. However, a huge amount of speaker information in SSL representations degrades timbre similarity and the quality of converted speech significantly. To address this problem, we proposed a high-similarity any-to-one voice conversion method with the input of SSL representations. We incorporated adversarial training mechanisms in the synthesis module using external unannotated corpora. Two auxiliary discriminators were trained to distinguish whether a sequence of mel-spectrograms has been converted by the acoustic model and whether a sequence of content embeddings contains speaker information from external corpora. Experimental results show that our proposed method achieves comparable similarity and higher naturalness than the supervised method, which needs a huge amount of annotated corpora for training and is applicable to improve similarity for VC methods with other SSL representations as input.
The recently-developed infant wearable MAIJU provides a means to automatically evaluate infants' motor performance in an objective and scalable manner in out-of-hospital settings. This information could be used for developmental research and to support clinical decision-making, such as detection of developmental problems and guiding of their therapeutic interventions. MAIJU-based analyses rely fully on the classification of infant's posture and movement; it is hence essential to study ways to increase the accuracy of such classifications, aiming to increase the reliability and robustness of the automated analysis. Here, we investigated how self-supervised pre-training improves performance of the classifiers used for analyzing MAIJU recordings, and we studied whether performance of the classifier models is affected by context-selective quality-screening of pre-training data to exclude periods of little infant movement or with missing sensors. Our experiments show that i) pre-training the classifier with unlabeled data leads to a robust accuracy increase of subsequent classification models, and ii) selecting context-relevant pre-training data leads to substantial further improvements in the classifier performance.