Recently, generative graph models have shown promising results in learning graph representations through self-supervised methods. However, most existing generative graph representation learning (GRL) approaches rely on random masking across the entire graph, which overlooks the entanglement of learned representations. This oversight results in non-robustness and a lack of explainability. Furthermore, disentangling the learned representations remains a significant challenge and has not been sufficiently explored in GRL research. Based on these insights, this paper introduces DiGGR (Disentangled Generative Graph Representation Learning), a self-supervised learning framework. DiGGR aims to learn latent disentangled factors and utilizes them to guide graph mask modeling, thereby enhancing the disentanglement of learned representations and enabling end-to-end joint learning. Extensive experiments on 11 public datasets for two different graph learning tasks demonstrate that DiGGR consistently outperforms many previous self-supervised methods, verifying the effectiveness of the proposed approach.
Recently, BigVGAN has emerged as high-performance speech vocoder. Its sequence-to-sequence-based synthesis, however, prohibits usage in low-latency conversational applications. Our work addresses this shortcoming in three steps. First, we introduce low latency into BigVGAN via implementing causal convolutions, yielding decreased performance. Second, to regain performance, we propose a teacher-student transfer learning scheme to distill the high-delay non-causal BigVGAN into our low-latency causal vocoder. Third, taking advantage of a self-supervised learning (SSL) model, in our case wav2vec 2.0, we align its encoder speech representations extracted from our low-latency causal vocoder to the ground truth ones. In speaker-independent settings, both proposed training schemes notably elevate the performance of our low-latency vocoder, closing up to the original high-delay BigVGAN. At only 21% higher complexity, our best small causal vocoder achieves 3.96 PESQ and 1.25 MCD, excelling even the original small non-causal BigVGAN (3.64 PESQ) by 0.32 PESQ and 0.1 MCD points, respectively.
Magnetic resonance imaging (MRI) is crucial in diagnosing various abdominal conditions and anomalies. Traditional MRI scans often yield anisotropic data due to technical constraints, resulting in varying resolutions across spatial dimensions, which limits diagnostic accuracy and volumetric analysis. Super-resolution (SR) techniques aim to address these limitations by reconstructing isotropic high-resolution images from anisotropic data. However, current SR methods often depend on indirect mappings and scarce 3D isotropic data for training, primarily focusing on two-dimensional enhancements rather than achieving genuine three-dimensional isotropy. We introduce ``SIMPLE,'' a Simultaneous Multi-Plane Self-Supervised Learning approach for isotropic MRI restoration from anisotropic data. Our method leverages existing anisotropic clinical data acquired in different planes, bypassing the need for simulated downsampling processes. By considering the inherent three-dimensional nature of MRI data, SIMPLE ensures realistic isotropic data generation rather than solely improving through-plane slices. This approach's flexibility allows it to be extended to multiple contrast types and acquisition methods commonly used in clinical settings. Our experiments on two distinct datasets (brain and abdomen) show that SIMPLE outperforms state-of-the-art methods both quantitatively using the Kernel Inception Distance (KID), semi-quantitatively through radiologist evaluations, and qualitatively through Fourier domain analysis. The generated isotropic volume facilitates more accurate volumetric analysis and 3D reconstructions, promising significant improvements in clinical diagnostic capabilities.
Prompting has become a practical method for utilizing pre-trained language models (LMs). This approach offers several advantages. It allows an LM to adapt to new tasks with minimal training and parameter updates, thus achieving efficiency in both storage and computation. Additionally, prompting modifies only the LM's inputs and harnesses the generative capabilities of language models to address various downstream tasks in a unified manner. This significantly reduces the need for human labor in designing task-specific models. These advantages become even more evident as the number of tasks served by the LM scales up. Motivated by the strengths of prompting, we are the first to explore the potential of prompting speech LMs in the domain of speech processing. Recently, there has been a growing interest in converting speech into discrete units for language modeling. Our pioneer research demonstrates that these quantized speech units are highly versatile within our unified prompting framework. Not only can they serve as class labels, but they also contain rich phonetic information that can be re-synthesized back into speech signals for speech generation tasks. Specifically, we reformulate speech processing tasks into speech-to-unit generation tasks. As a result, we can seamlessly integrate tasks such as speech classification, sequence generation, and speech generation within a single, unified prompting framework. The experiment results show that the prompting method can achieve competitive performance compared to the strong fine-tuning method based on self-supervised learning models with a similar number of trainable parameters. The prompting method also shows promising results in the few-shot setting. Moreover, with the advanced speech LMs coming into the stage, the proposed prompting framework attains great potential.
Masked Image Modeling (MIM) is a technique in self-supervised learning that focuses on acquiring detailed visual representations from unlabeled images by estimating the missing pixels in randomly masked sections. It has proven to be a powerful tool for the preliminary training of Vision Transformers (ViTs), yielding impressive results across various tasks. Nevertheless, most MIM methods heavily depend on the random masking strategy to formulate the pretext task. This strategy necessitates numerous trials to ascertain the optimal dropping ratio, which can be resource-intensive, requiring the model to be pre-trained for anywhere between 800 to 1600 epochs. Furthermore, this approach may not be suitable for all datasets. In this work, we propose a new masking strategy that effectively helps the model capture global and local features. Based on this masking strategy, SymMIM, our proposed training pipeline for MIM is introduced. SymMIM achieves a new SOTA accuracy of 85.9\% on ImageNet using ViT-Large and surpasses previous SOTA across downstream tasks such as image classification, semantic segmentation, object detection, instance segmentation tasks, and so on.
Lagrangian decomposition (LD) is a relaxation method that provides a dual bound for constrained optimization problems by decomposing them into more manageable sub-problems. This bound can be used in branch-and-bound algorithms to prune the search space effectively. In brief, a vector of Lagrangian multipliers is associated with each sub-problem, and an iterative procedure (e.g., a sub-gradient optimization) adjusts these multipliers to find the tightest bound. Initially applied to integer programming, Lagrangian decomposition also had success in constraint programming due to its versatility and the fact that global constraints provide natural sub-problems. However, the non-linear and combinatorial nature of sub-problems in constraint programming makes it computationally intensive to optimize the Lagrangian multipliers with sub-gradient methods at each node of the tree search. This currently limits the practicality of LD as a general bounding mechanism for constraint programming. To address this challenge, we propose a self-supervised learning approach that leverages neural networks to generate multipliers directly, yielding tight bounds. This approach significantly reduces the number of sub-gradient optimization steps required, enhancing the pruning efficiency and reducing the execution time of constraint programming solvers. This contribution is one of the few that leverage learning to enhance bounding mechanisms on the dual side, a critical element in the design of combinatorial solvers. To our knowledge, this work presents the first generic method for learning valid dual bounds in constraint programming.
Speech representation models based on the transformer architecture and trained by self-supervised learning have shown great promise for solving tasks such as speech and speaker recognition, keyword spotting, emotion detection, and more. Typically, it is found that larger models lead to better performance. However, the significant computational effort involved in such large transformer systems is a challenge for embedded and real-world applications. Recent work has shown that there is significant redundancy in the transformer models for NLP and massive layer pruning is feasible (Sajjad et al., 2023). Here, we investigate layer pruning in audio models. We base the pruning decision on a convexity criterion. Convexity of classification regions has recently been proposed as an indicator of subsequent fine-tuning performance in a range of application domains, including NLP and audio. In empirical investigations, we find a massive reduction in the computational effort with no loss of performance or even improvements in certain cases.
Foley sound synthesis is crucial for multimedia production, enhancing user experience by synchronizing audio and video both temporally and semantically. Recent studies on automating this labor-intensive process through video-to-sound generation face significant challenges. Systems lacking explicit temporal features suffer from poor alignment and controllability, while timestamp-based models require costly and subjective human annotation. We propose Video-Foley, a video-to-sound system using Root Mean Square (RMS) as an intuitive condition with semantic timbre prompts (audio or text). RMS, a frame-level intensity envelope closely related to audio semantics, acts as a temporal event feature to guide audio generation from video. The annotation-free self-supervised learning framework consists of two stages, Video2RMS and RMS2Sound, incorporating novel ideas including RMS discretization and RMS-ControlNet with a pretrained text-to-audio model. Our extensive evaluation shows that Video-Foley achieves state-of-the-art performance in audio-visual alignment and controllability for sound timing, intensity, timbre, and nuance. Source code, model weights and demos are available on our companion website. (https://jnwnlee.github.io/video-foley-demo)
Recent advances in SSL enabled novel medical AI models, known as foundation models, offer great potential for better characterizing health from diverse biomedical data. CGM provides rich, temporal data on glycemic patterns, but its full potential for predicting broader health outcomes remains underutilized. Here, we present GluFormer, a generative foundation model for CGM data that learns nuanced glycemic patterns and translates them into predictive representations of metabolic health. Trained on over 10 million CGM measurements from 10,812 adults, primarily without diabetes, GluFormer uses autoregressive token prediction to capture longitudinal glucose dynamics. We show that GluFormer generalizes to 19 external cohorts (n=6,044) spanning different ethnicities and ages, 5 countries, 8 CGM devices, and diverse pathophysiological states. GluFormers representations exceed the performance of current CGM metrics, such as the Glucose Management Indicator (GMI), for forecasting clinical measures. In a longitudinal study of 580 adults with CGM data and 12-year follow-up, GluFormer identifies individuals at elevated risk of developing diabetes more effectively than blood HbA1C%, capturing 66% of all new-onset diabetes diagnoses in the top quartile versus 7% in the bottom quartile. Similarly, 69% of cardiovascular-death events occurred in the top quartile with none in the bottom quartile, demonstrating powerful risk stratification beyond traditional glycemic metrics. We also show that CGM representations from pre-intervention periods in Randomized Clinical Trials outperform other methods in predicting primary and secondary outcomes. When integrating dietary data into GluFormer, we show that the multi-modal version of the model can accurately generate CGM data based on dietary intake data, simulate outcomes of dietary interventions, and predict individual responses to specific foods.
The potential of deep learning in clinical speech processing is immense, yet the hurdles of limited and imbalanced clinical data samples loom large. This article addresses these challenges by showcasing the utilization of automatic speech recognition and self-supervised learning representations, pre-trained on extensive datasets of normal speech. This innovative approach aims to estimate voice quality of patients with impaired vocal systems. Experiments involve checks on PVQD dataset, covering various causes of vocal system damage in English, and a Japanese dataset focusing on patients with Parkinson's disease before and after undergoing subthalamic nucleus deep brain stimulation (STN-DBS) surgery. The results on PVQD reveal a notable correlation (>0.8 on PCC) and an extraordinary accuracy (<0.5 on MSE) in predicting Grade, Breathy, and Asthenic indicators. Meanwhile, progress has been achieved in predicting the voice quality of patients in the context of STN-DBS.
The proliferation of geospatial data in urban and territorial environments has significantly facilitated the development of geospatial artificial intelligence (GeoAI) across various urban applications. Given the vast yet inherently sparse labeled nature of geospatial data, there is a critical need for techniques that can effectively leverage such data without heavy reliance on labeled datasets. This requirement aligns with the principles of self-supervised learning (SSL), which has attracted increasing attention for its adoption in geospatial data. This paper conducts a comprehensive and up-to-date survey of SSL techniques applied to or developed for three primary data (geometric) types prevalent in geospatial vector data: points, polylines, and polygons. We systematically categorize various SSL techniques into predictive and contrastive methods, discussing their application with respect to each data type in enhancing generalization across various downstream tasks. Furthermore, we review the emerging trends of SSL for GeoAI, and several task-specific SSL techniques. Finally, we discuss several key challenges in the current research and outline promising directions for future investigation. By presenting a structured analysis of relevant studies, this paper aims to inspire continued advancements in the integration of SSL with GeoAI, encouraging innovative methods to harnessing the power of geospatial data.
Transcriptome foundation models TFMs hold great promises of deciphering the transcriptomic language that dictate diverse cell functions by self-supervised learning on large-scale single-cell gene expression data, and ultimately unraveling the complex mechanisms of human diseases. However, current TFMs treat cells as independent samples and ignore the taxonomic relationships between cell types, which are available in cell ontology graphs. We argue that effectively leveraging this ontology information during the TFM pre-training can improve learning biologically meaningful gene co-expression patterns while preserving TFM as a general purpose foundation model for downstream zero-shot and fine-tuning tasks. To this end, we present \textbf{s}ingle \textbf{c}ell, \textbf{Cell}-\textbf{o}ntology guided TFM scCello. We introduce cell-type coherence loss and ontology alignment loss, which are minimized along with the masked gene expression prediction loss during the pre-training. The novel loss component guide scCello to learn the cell-type-specific representation and the structural relation between cell types from the cell ontology graph, respectively. We pre-trained scCello on 22 million cells from CellxGene database leveraging their cell-type labels mapped to the cell ontology graph from Open Biological and Biomedical Ontology Foundry. Our TFM demonstrates competitive generalization and transferability performance over the existing TFMs on biologically important tasks including identifying novel cell types of unseen cells, prediction of cell-type-specific marker genes, and cancer drug responses.
Modern consumer cameras commonly employ the rolling shutter (RS) imaging mechanism, via which images are captured by scanning scenes row-by-row, resulting in RS distortion for dynamic scenes. To correct RS distortion, existing methods adopt a fully supervised learning manner that requires high framerate global shutter (GS) images as ground-truth for supervision. In this paper, we propose an enhanced Self-supervised learning framework for Dual reversed RS distortion Correction (SelfDRSC++). Firstly, we introduce a lightweight DRSC network that incorporates a bidirectional correlation matching block to refine the joint optimization of optical flows and corrected RS features, thereby improving correction performance while reducing network parameters. Subsequently, to effectively train the DRSC network, we propose a self-supervised learning strategy that ensures cycle consistency between input and reconstructed dual reversed RS images. The RS reconstruction in SelfDRSC++ can be interestingly formulated as a specialized instance of video frame interpolation, where each row in reconstructed RS images is interpolated from predicted GS images by utilizing RS distortion time maps. By achieving superior performance while simplifying the training process, SelfDRSC++ enables feasible one-stage self-supervised training. Additionally, besides start and end RS scanning time, SelfDRSC++ allows supervision of GS images at arbitrary intermediate scanning times, thus enabling the learned DRSC network to generate high framerate GS videos. The code and trained models are available at \url{https://github.com/shangwei5/SelfDRSC_plusplus}.
Existing Vehicle-to-Everything (V2X) cooperative perception methods rely on accurate multi-agent 3D annotations. Nevertheless, it is time-consuming and expensive to collect and annotate real-world data, especially for V2X systems. In this paper, we present a self-supervised learning method for V2X cooperative perception, which utilizes the vast amount of unlabeled 3D V2X data to enhance the perception performance. Beyond simply extending the previous pre-training methods for point-cloud representation learning, we introduce a novel self-supervised Cooperative Pretraining framework (termed as CooPre) customized for a collaborative scenario. We point out that cooperative point-cloud sensing compensates for information loss among agents. This motivates us to design a novel proxy task for the 3D encoder to reconstruct LiDAR point clouds across different agents. Besides, we develop a V2X bird-eye-view (BEV) guided masking strategy which effectively allows the model to pay attention to 3D features across heterogeneous V2X agents (i.e., vehicles and infrastructure) in the BEV space. Noticeably, such a masking strategy effectively pretrains the 3D encoder and is compatible with mainstream cooperative perception backbones. Our approach, validated through extensive experiments on representative datasets (i.e., V2X-Real, V2V4Real, and OPV2V), leads to a performance boost across all V2X settings. Additionally, we demonstrate the framework's improvements in cross-domain transferability, data efficiency, and robustness under challenging scenarios. The code will be made publicly available.
Self-supervised learning has driven significant progress in learning from single-subject, iconic images. However, there are still unanswered questions about the use of minimally-curated, naturalistic video data, which contain dense scenes with many independent objects, imbalanced class distributions, and varying object sizes. In this paper, we propose PooDLe, a self-supervised learning method that combines an invariance-based objective on pooled representations with a dense SSL objective that enforces equivariance to optical flow warping. Our results show that a unified objective applied at multiple feature scales is essential for learning effective image representations from naturalistic videos. We validate our method with experiments on the BDD100K driving video dataset and the Walking Tours first-person video dataset, demonstrating its ability to capture spatial understanding from a dense objective and semantic understanding via a pooled representation objective.
Self-supervised learning~(SSL) is essential to obtain foundation models in NLP and CV domains via effectively leveraging knowledge in large-scale unlabeled data. The reason for its success is that a suitable SSL design can help the model to follow the neural scaling law, i.e., the performance consistently improves with increasing model and dataset sizes. However, it remains a mystery whether existing SSL in the graph domain can follow the scaling behavior toward building Graph Foundation Models~(GFMs) with large-scale pre-training. In this study, we examine whether existing graph SSL techniques can follow the neural scaling behavior with the potential to serve as the essential component for GFMs. Our benchmark includes comprehensive SSL technique implementations with analysis conducted on both the conventional SSL setting and many new settings adopted in other domains. Surprisingly, despite the SSL loss continuously decreasing, no existing graph SSL techniques follow the neural scaling behavior on the downstream performance. The model performance only merely fluctuates on different data scales and model scales. Instead of the scales, the key factors influencing the performance are the choices of model architecture and pretext task design. This paper examines existing SSL techniques for the feasibility of Graph SSL techniques in developing GFMs and opens a new direction for graph SSL design with the new evaluation prototype. Our code implementation is available online to ease reproducibility on https://github.com/GraphSSLScaling/GraphSSLScaling.
We present OCTCube-M, a 3D OCT-based multi-modal foundation model for jointly analyzing OCT and en face images. OCTCube-M first developed OCTCube, a 3D foundation model pre-trained on 26,685 3D OCT volumes encompassing 1.62 million 2D OCT images. It then exploits a novel multi-modal contrastive learning framework COEP to integrate other retinal imaging modalities, such as fundus autofluorescence and infrared retinal imaging, into OCTCube, efficiently extending it into multi-modal foundation models. OCTCube achieves best performance on predicting 8 retinal diseases, demonstrating strong generalizability on cross-cohort, cross-device and cross-modality prediction. OCTCube can also predict cross-organ nodule malignancy (CT) and low cardiac ejection fraction as well as systemic diseases, such as diabetes and hypertension, revealing its wide applicability beyond retinal diseases. We further develop OCTCube-IR using COEP with 26,685 OCT and IR image pairs. OCTCube-IR can accurately retrieve between OCT and IR images, allowing joint analysis between 3D and 2D retinal imaging modalities. Finally, we trained a tri-modal foundation model OCTCube-EF from 4 million 2D OCT images and 400K en face retinal images. OCTCube-EF attains the best performance on predicting the growth rate of geographic atrophy (GA) across datasets collected from 6 multi-center global trials conducted in 23 countries. This improvement is statistically equivalent to running a clinical trial with more than double the size of the original study. Our analysis based on another retrospective case study reveals OCTCube-EF's ability to avoid false positive Phase-III results according to its accurate treatment effect estimation on the Phase-II results. In sum, OCTCube-M is a 3D multi-modal foundation model framework that integrates OCT and other retinal imaging modalities revealing substantial diagnostic and prognostic benefits.
Speech modeling methods learn one embedding for a fixed segment of speech, typically in between 10-25 ms. The information present in speech can be divided into two categories: "what is being said" (content) and "how it is expressed" (other) and these two are orthogonal in nature causing the optimization algorithm to find a sub-optimal solution if forced to optimize together. This leads to sub-optimal performance in one or all downstream tasks as shown by previous studies. Current self-supervised learning (SSL) methods such as HuBERT are very good at modeling the content information present in speech. Data augmentation improves the performance on tasks which require effective modeling of other information but this leads to a divided capacity of the model. In this work, we conduct a preliminary study to understand the importance of modeling other information using separate learnable parameters. We propose a modified version of HuBERT, termed Other HuBERT (O-HuBERT), to test our hypothesis. Our findings are twofold: first, the O-HuBERT method is able to utilize all layers to build complex features to encode other information; second, a robust data augmentation strategy is essential for learning the information required by tasks that depend on other information and to achieve state-of-the-art (SOTA) performance on the SUPERB benchmark with a similarly sized model (100 million parameters) and pre-training data (960 hours).
Contrastive representation learning, which aims to learnthe shared information between different views of unlabeled data by maximizing the mutual information between them, has shown its powerful competence in self-supervised learning for downstream tasks. However, recent works have demonstrated that more estimated mutual information does not guarantee better performance in different downstream tasks. Such works inspire us to conjecture that the learned representations not only maintain task-relevant information from unlabeled data but also carry task-irrelevant information which is superfluous for downstream tasks, thus leading to performance degeneration. In this paper we show that superfluous information does exist during the conventional contrastive learning framework, and further design a new objective, namely SuperInfo, to learn robust representations by a linear combination of both predictive and superfluous information. Besides, we notice that it is feasible to tune the coefficients of introduced losses to discard task-irrelevant information, while keeping partial non-shared task-relevant information according to our SuperInfo loss.We demonstrate that learning with our loss can often outperform the traditional contrastive learning approaches on image classification, object detection and instance segmentation tasks with significant improvements.
While recent zero-shot multi-speaker text-to-speech (TTS) models achieve impressive results, they typically rely on extensive transcribed speech datasets from numerous speakers and intricate training pipelines. Meanwhile, self-supervised learning (SSL) speech features have emerged as effective intermediate representations for TTS. Further, SSL features from different speakers that are linearly close share phonetic information while maintaining individual speaker identity. In this study, we introduce kNN-TTS, a simple and effective framework for zero-shot multi-speaker TTS using retrieval methods which leverage the linear relationships between SSL features. Objective and subjective evaluations show that our models, trained on transcribed speech from a single speaker only, achieve performance comparable to state-of-the-art models that are trained on significantly larger training datasets. The low training data requirements mean that kNN-TTS is well suited for the development of multi-speaker TTS systems for low-resource domains and languages. We also introduce an interpolation parameter which enables fine-grained voice morphing. Demo samples are available at https://idiap.github.io/knn-tts