Recent generative models produce images with a level of authenticity that makes them nearly indistinguishable from real photos and artwork. Potential harmful use cases of these models, necessitate the creation of robust synthetic image detectors. However, current datasets in the field contain generated images with questionable quality or have examples from one predominant content type which leads to poor generalizability of the underlying detectors. We find that the curation of a balanced amount of high-resolution generated images across various content types is crucial for the generalizability of detectors, and introduce ImagiNet, a dataset of 200K examples, spanning four categories: photos, paintings, faces, and miscellaneous. Synthetic images in ImagiNet are produced with both open-source and proprietary generators, whereas real counterparts for each content type are collected from public datasets. The structure of ImagiNet allows for a two-track evaluation system: i) classification as real or synthetic and ii) identification of the generative model. To establish a strong baseline, we train a ResNet-50 model using a self-supervised contrastive objective (SelfCon) for each track which achieves evaluation AUC of up to 0.99 and balanced accuracy ranging from 86% to 95%, even under conditions that involve compression and resizing. The provided model is generalizable enough to achieve zero-shot state-of-the-art performance on previous synthetic detection benchmarks. We provide ablations to demonstrate the importance of content types and publish code and data.
Surgical instrument segmentation is crucial in surgical scene understanding, thereby facilitating surgical safety. Existing algorithms directly detected all instruments of pre-defined categories in the input image, lacking the capability to segment specific instruments according to the surgeon's intention. During different stages of surgery, surgeons exhibit varying preferences and focus toward different surgical instruments. Therefore, an instrument segmentation algorithm that adheres to the surgeon's intention can minimize distractions from irrelevant instruments and assist surgeons to a great extent. The recent Segment Anything Model (SAM) reveals the capability to segment objects following prompts, but the manual annotations for prompts are impractical during the surgery. To address these limitations in operating rooms, we propose an audio-driven surgical instrument segmentation framework, named ASI-Seg, to accurately segment the required surgical instruments by parsing the audio commands of surgeons. Specifically, we propose an intention-oriented multimodal fusion to interpret the segmentation intention from audio commands and retrieve relevant instrument details to facilitate segmentation. Moreover, to guide our ASI-Seg segment of the required surgical instruments, we devise a contrastive learning prompt encoder to effectively distinguish the required instruments from the irrelevant ones. Therefore, our ASI-Seg promotes the workflow in the operating rooms, thereby providing targeted support and reducing the cognitive load on surgeons. Extensive experiments are performed to validate the ASI-Seg framework, which reveals remarkable advantages over classical state-of-the-art and medical SAMs in both semantic segmentation and intention-oriented segmentation. The source code is available at https://github.com/Zonmgin-Zhang/ASI-Seg.
The burgeoning short video industry has accelerated the advancement of video-music retrieval technology, assisting content creators in selecting appropriate music for their videos. In self-supervised training for video-to-music retrieval, the video and music samples in the dataset are separated from the same video work, so they are all one-to-one matches. This does not match the real situation. In reality, a video can use different music as background music, and a music can be used as background music for different videos. Many videos and music that are not in a pair may be compatible, leading to false negative noise in the dataset. A novel inter-intra modal (II) loss is proposed as a solution. By reducing the variation of feature distribution within the two modalities before and after the encoder, II loss can reduce the model's overfitting to such noise without removing it in a costly and laborious way. The video-music retrieval framework, II-CLVM (Contrastive Learning for Video-Music Retrieval), incorporating the II Loss, achieves state-of-the-art performance on the YouTube8M dataset. The framework II-CLVTM shows better performance when retrieving music using multi-modal video information (such as text in videos). Experiments are designed to show that II loss can effectively alleviate the problem of false negative noise in retrieval tasks. Experiments also show that II loss improves various self-supervised and supervised uni-modal and cross-modal retrieval tasks, and can obtain good retrieval models with a small amount of training samples.
Proteins govern most biological functions essential for life, but achieving controllable protein discovery and optimization remains challenging. Recently, machine learning-assisted protein editing (MLPE) has shown promise in accelerating optimization cycles and reducing experimental workloads. However, current methods struggle with the vast combinatorial space of potential protein edits and cannot explicitly conduct protein editing using biotext instructions, limiting their interactivity with human feedback. To fill these gaps, we propose a novel method called ProtET for efficient CLIP-informed protein editing through multi-modality learning. Our approach comprises two stages: in the pretraining stage, contrastive learning aligns protein-biotext representations encoded by two large language models (LLMs), respectively. Subsequently, during the protein editing stage, the fused features from editing instruction texts and original protein sequences serve as the final editing condition for generating target protein sequences. Comprehensive experiments demonstrated the superiority of ProtET in editing proteins to enhance human-expected functionality across multiple attribute domains, including enzyme catalytic activity, protein stability and antibody specific binding ability. And ProtET improves the state-of-the-art results by a large margin, leading to significant stability improvements of 16.67% and 16.90%. This capability positions ProtET to advance real-world artificial protein editing, potentially addressing unmet academic, industrial, and clinical needs.
We introduce a novel self-supervised deep clustering approach tailored for unstructured data without requiring prior knowledge of the number of clusters, termed Adaptive Self-supervised Robust Clustering (ASRC). In particular, ASRC adaptively learns the graph structure and edge weights to capture both local and global structural information. The obtained graph enables us to learn clustering-friendly feature representations by an enhanced graph auto-encoder with contrastive learning technique. It further leverages the clustering results adaptively obtained by robust continuous clustering (RCC) to generate prototypes for negative sampling, which can further contribute to promoting consistency among positive pairs and enlarging the gap between positive and negative samples. ASRC obtains the final clustering results by applying RCC to the learned feature representations with their consistent graph structure and edge weights. Extensive experiments conducted on seven benchmark datasets demonstrate the efficacy of ASRC, demonstrating its superior performance over other popular clustering models. Notably, ASRC even outperforms methods that rely on prior knowledge of the number of clusters, highlighting its effectiveness in addressing the challenges of clustering unstructured data.
Virtual screening (VS) is a critical step in computer-aided drug discovery, aiming to identify molecules that bind to a specific target receptor like protein. Traditional VS methods, such as docking, are often too time-consuming for screening large-scale molecular databases. Recent advances in deep learning have demonstrated that learning vector representations for both proteins and molecules using contrastive learning can outperform traditional docking methods. However, given that target databases often contain billions of molecules, real-valued vector representations adopted by existing methods can still incur significant memory and time costs in VS. To address this problem, in this paper we propose a hashing-based contrastive learning method, called DrugHash, for VS. DrugHash treats VS as a retrieval task that uses efficient binary hash codes for retrieval. In particular, DrugHash designs a simple yet effective hashing strategy to enable end-to-end learning of binary hash codes for both protein and molecule modalities, which can dramatically reduce the memory and time costs with higher accuracy compared with existing methods. Experimental results show that DrugHash can outperform existing methods to achieve state-of-the-art accuracy, with a memory saving of 32$\times$ and a speed improvement of 3.5$\times$.
Transcription-only Supervised Text Spotting aims to learn text spotters relying only on transcriptions but no text boundaries for supervision, thus eliminating expensive boundary annotation. The crux of this task lies in locating each transcription in scene text images without location annotations. In this work, we formulate this challenging problem as a Weakly Supervised Cross-modality Contrastive Learning problem, and design a simple yet effective model dubbed WeCromCL that is able to detect each transcription in a scene image in a weakly supervised manner. Unlike typical methods for cross-modality contrastive learning that focus on modeling the holistic semantic correlation between an entire image and a text description, our WeCromCL conducts atomistic contrastive learning to model the character-wise appearance consistency between a text transcription and its correlated region in a scene image to detect an anchor point for the transcription in a weakly supervised manner. The detected anchor points by WeCromCL are further used as pseudo location labels to guide the learning of text spotting. Extensive experiments on four challenging benchmarks demonstrate the superior performance of our model over other methods. Code will be released.
Dysarthric speech recognition (DSR) presents a formidable challenge due to inherent inter-speaker variability, leading to severe performance degradation when applying DSR models to new dysarthric speakers. Traditional speaker adaptation methodologies typically involve fine-tuning models for each speaker, but this strategy is cost-prohibitive and inconvenient for disabled users, requiring substantial data collection. To address this issue, we introduce a prototype-based approach that markedly improves DSR performance for unseen dysarthric speakers without additional fine-tuning. Our method employs a feature extractor trained with HuBERT to produce per-word prototypes that encapsulate the characteristics of previously unseen speakers. These prototypes serve as the basis for classification. Additionally, we incorporate supervised contrastive learning to refine feature extraction. By enhancing representation quality, we further improve DSR performance, enabling effective personalized DSR. We release our code at https://github.com/NKU-HLT/PB-DSR.
Brain network analysis plays an increasingly important role in studying brain function and the exploring of disease mechanisms. However, existing brain network construction tools have some limitations, including dependency on empirical users, weak consistency in repeated experiments and time-consuming processes. In this work, a diffusion-based brain network pipeline, DGCL is designed for end-to-end construction of brain networks. Initially, the brain region-aware module (BRAM) precisely determines the spatial locations of brain regions by the diffusion process, avoiding subjective parameter selection. Subsequently, DGCL employs graph contrastive learning to optimize brain connections by eliminating individual differences in redundant connections unrelated to diseases, thereby enhancing the consistency of brain networks within the same group. Finally, the node-graph contrastive loss and classification loss jointly constrain the learning process of the model to obtain the reconstructed brain network, which is then used to analyze important brain connections. Validation on two datasets, ADNI and ABIDE, demonstrates that DGCL surpasses traditional methods and other deep learning models in predicting disease development stages. Significantly, the proposed model improves the efficiency and generalization of brain network construction. In summary, the proposed DGCL can be served as a universal brain network construction scheme, which can effectively identify important brain connections through generative paradigms and has the potential to provide disease interpretability support for neuroscience research.
Representation learning has emerged as a powerful paradigm for extracting valuable latent features from complex, high-dimensional data. In financial domains, learning informative representations for assets can be used for tasks like sector classification, and risk management. However, the complex and stochastic nature of financial markets poses unique challenges. We propose a novel contrastive learning framework to generate asset embeddings from financial time series data. Our approach leverages the similarity of asset returns over many subwindows to generate informative positive and negative samples, using a statistical sampling strategy based on hypothesis testing to address the noisy nature of financial data. We explore various contrastive loss functions that capture the relationships between assets in different ways to learn a discriminative representation space. Experiments on real-world datasets demonstrate the effectiveness of the learned asset embeddings on benchmark industry classification and portfolio optimization tasks. In each case our novel approaches significantly outperform existing baselines highlighting the potential for contrastive learning to capture meaningful and actionable relationships in financial data.
Tracking in gigapixel scenarios holds numerous potential applications in video surveillance and pedestrian analysis. Existing algorithms attempt to perform tracking in crowded scenes by utilizing multiple cameras or group relationships. However, their performance significantly degrades when confronted with complex interaction and occlusion inherent in gigapixel images. In this paper, we introduce DynamicTrack, a dynamic tracking framework designed to address gigapixel tracking challenges in crowded scenes. In particular, we propose a dynamic detector that utilizes contrastive learning to jointly detect the head and body of pedestrians. Building upon this, we design a dynamic association algorithm that effectively utilizes head and body information for matching purposes. Extensive experiments show that our tracker achieves state-of-the-art performance on widely used tracking benchmarks specifically designed for gigapixel crowded scenes.
Recently, large-scale visual language pre-trained (VLP) models have demonstrated impressive performance across various downstream tasks. Motivated by these advancements, pioneering efforts have emerged in multi-label image recognition with missing labels, leveraging VLP prompt-tuning technology. However, they usually cannot match text and vision features well, due to complicated semantics gaps and missing labels in a multi-label image. To tackle this challenge, we propose $\textbf{T}$ext-$\textbf{R}$egion $\textbf{M}$atching for optimizing $\textbf{M}$ulti-$\textbf{L}$abel prompt tuning, namely TRM-ML, a novel method for enhancing meaningful cross-modal matching. Compared to existing methods, we advocate exploring the information of category-aware regions rather than the entire image or pixels, which contributes to bridging the semantic gap between textual and visual representations in a one-to-one matching manner. Concurrently, we further introduce multimodal contrastive learning to narrow the semantic gap between textual and visual modalities and establish intra-class and inter-class relationships. Additionally, to deal with missing labels, we propose a multimodal category prototype that leverages intra- and inter-category semantic relationships to estimate unknown labels, facilitating pseudo-label generation. Extensive experiments on the MS-COCO, PASCAL VOC, Visual Genome, NUS-WIDE, and CUB-200-211 benchmark datasets demonstrate that our proposed framework outperforms the state-of-the-art methods by a significant margin. Our code is available here: https://github.com/yu-gi-oh-leilei/TRM-ML.
Brain-computer interfaces (BCIs) offer a means to convert neural signals into control signals, providing a potential restoration of movement for people with paralysis. Despite their promise, BCIs face a significant challenge in maintaining decoding accuracy over time due to neural nonstationarities. However, the decoding accuracy of BCI drops severely across days due to the neural data drift. While current recalibration techniques address this issue to a degree, they often fail to leverage the limited labeled data, to consider the signal correlation between two days, or to perform conditional alignment in regression tasks. This paper introduces a novel approach to enhance recalibration performance. We begin with preliminary experiments that reveal the temporal patterns of neural signal changes and identify three critical elements for effective recalibration: global alignment, conditional speed alignment, and feature-label consistency. Building on these insights, we propose the Speed-enhanced Subdomain Adaptation Regression (SSAR) framework, integrating semi-supervised learning with domain adaptation techniques in regression neural decoding. SSAR employs Speed-enhanced Subdomain Alignment (SeSA) for global and speed conditional alignment of similarly labeled data, with Contrastive Consistency Constraint (CCC) to enhance the alignment of SeSA by reinforcing feature-label consistency through contrastive learning. Our comprehensive set of experiments, both qualitative and quantitative, substantiate the superior recalibration performance and robustness of SSAR.
We present Banyan, a model that efficiently learns semantic representations by leveraging explicit hierarchical structure. While transformers excel at scale, they struggle in low-resource settings. Conversely recent structured models have shown promise as efficient learners, but lack performance. Banyan bridges this gap with two key innovations: an entangled hierarchical tree structure and diagonalized message passing, enabling it to outperform larger transformer models with just 14 non-embedding parameters. It excels in low-resource settings, offering a viable alternative for under-represented languages and highlighting its potential for efficient, interpretable NLP in resource-constrained environments.
Graph recommender (GR) is a type of graph neural network (GNNs) encoder that is customized for extracting information from the user-item interaction graph. Due to its strong performance on the recommendation task, GR has gained significant attention recently. Graph contrastive learning (GCL) is also a popular research direction that aims to learn, often unsupervised, GNNs with certain contrastive objectives. As a general graph representation learning method, GCLs have been widely adopted with the supervised recommendation loss for joint training of GRs. Despite the intersection of GR and GCL research, theoretical understanding of the relationship between the two fields is surprisingly sparse. This vacancy inevitably leads to inefficient scientific research. In this paper, we aim to bridge the gap between the field of GR and GCL from the perspective of encoders and loss functions. With mild assumptions, we theoretically show an astonishing fact that graph recommender is equivalent to a commonly-used single-view graph contrastive model. Specifically, we find that (1) the classic encoder in GR is essentially a linear graph convolutional network with one-hot inputs, and (2) the loss function in GR is well bounded by a single-view GCL loss with certain hyperparameters. The first observation enables us to explain crucial designs of GR models, e.g., the removal of self-loop and nonlinearity. And the second finding can easily prompt many cross-field research directions. We empirically show a remarkable result that the recommendation loss and the GCL loss can be used interchangeably. The fact that we can train GR models solely with the GCL loss is particularly insightful, since before this work, GCLs were typically viewed as unsupervised methods that need fine-tuning. We also discuss some potential future works inspired by our theory.
Learning good representations involves capturing the diverse ways in which data samples relate. Contrastive loss - an objective matching related samples - underlies methods from self-supervised to multimodal learning. Contrastive losses, however, can be viewed more broadly as modifying a similarity graph to indicate how samples should relate in the embedding space. This view reveals a shortcoming in contrastive learning: the similarity graph is binary, as only one sample is the related positive sample. Crucially, similarities \textit{across} samples are ignored. Based on this observation, we revise the standard contrastive loss to explicitly encode how a sample relates to others. We experiment with this new objective, called $\mathbb{X}$-Sample Contrastive, to train vision models based on similarities in class or text caption descriptions. Our study spans three scales: ImageNet-1k with 1 million, CC3M with 3 million, and CC12M with 12 million samples. The representations learned via our objective outperform both contrastive self-supervised and vision-language models trained on the same data across a range of tasks. When training on CC12M, we outperform CLIP by $0.6\%$ on both ImageNet and ImageNet Real. Our objective appears to work particularly well in lower-data regimes, with gains over CLIP of $16.8\%$ on ImageNet and $18.1\%$ on ImageNet Real when training with CC3M. Finally, our objective seems to encourage the model to learn representations that separate objects from their attributes and backgrounds, with gains of $3.3$-$5.6$\% over CLIP on ImageNet9. We hope the proposed solution takes a small step towards developing richer learning objectives for understanding sample relations in foundation models.
Predicting traffic accidents is the key to sustainable city management, which requires effective address of the dynamic and complex spatiotemporal characteristics of cities. Current data-driven models often struggle with data sparsity and typically overlook the integration of diverse urban data sources and the high-order dependencies within them. Additionally, they frequently rely on predefined topologies or weights, limiting their adaptability in spatiotemporal predictions. To address these issues, we introduce the Spatiotemporal Multiview Adaptive HyperGraph Learning (SMA-Hyper) model, a dynamic deep learning framework designed for traffic accident prediction. Building on previous research, this innovative model incorporates dual adaptive spatiotemporal graph learning mechanisms that enable high-order cross-regional learning through hypergraphs and dynamic adaptation to evolving urban data. It also utilises contrastive learning to enhance global and local data representations in sparse datasets and employs an advance attention mechanism to fuse multiple views of accident data and urban functional features, thereby enriching the contextual understanding of risk factors. Extensive testing on the London traffic accident dataset demonstrates that the SMA-Hyper model significantly outperforms baseline models across various temporal horizons and multistep outputs, affirming the effectiveness of its multiview fusion and adaptive learning strategies. The interpretability of the results further underscores its potential to improve urban traffic management and safety by leveraging complex spatiotemporal urban data, offering a scalable framework adaptable to diverse urban environments.
The rise of social media and the exponential growth of multimodal communication necessitates advanced techniques for Multimodal Information Extraction (MIE). However, existing methodologies primarily rely on direct Image-Text interactions, a paradigm that often faces significant challenges due to semantic and modality gaps between images and text. In this paper, we introduce a new paradigm of Image-Context-Text interaction, where large multimodal models (LMMs) are utilized to generate descriptive textual context to bridge these gaps. In line with this paradigm, we propose a novel Shapley Value-based Contrastive Alignment (Shap-CA) method, which aligns both context-text and context-image pairs. Shap-CA initially applies the Shapley value concept from cooperative game theory to assess the individual contribution of each element in the set of contexts, texts and images towards total semantic and modality overlaps. Following this quantitative evaluation, a contrastive learning strategy is employed to enhance the interactive contribution within context-text/image pairs, while minimizing the influence across these pairs. Furthermore, we design an adaptive fusion module for selective cross-modal fusion. Extensive experiments across four MIE datasets demonstrate that our method significantly outperforms existing state-of-the-art methods.
Contrastive Learning (CL)-based recommender systems have gained prominence in the context of Heterogeneous Graph (HG) due to their capacity to enhance the consistency of representations across different views. However, existing frameworks often neglect the fact that user-item interactions within HG are governed by diverse latent intents (e.g., brand preferences or demographic characteristics of item audiences), which are pivotal in capturing fine-grained relations. The exploration of these underlying intents, particularly through the lens of meta-paths in HGs, presents us with two principal challenges: i) How to integrate CL with intents; ii) How to mitigate noise from meta-path-driven intents. To address these challenges, we propose an innovative framework termed Intent-guided Heterogeneous Graph Contrastive Learning (IHGCL), which designed to enhance CL-based recommendation by capturing the intents contained within meta-paths. Specifically, the IHGCL framework includes: i) a meta-path-based Dual Contrastive Learning (DCL) approach to effectively integrate intents into the recommendation, constructing intent-intent contrast and intent-interaction contrast; ii) a Bottlenecked AutoEncoder (BAE) that combines mask propagation with the information bottleneck principle to significantly reduce noise perturbations introduced by meta-paths. Empirical evaluations conducted across six distinct datasets demonstrate the superior performance of our IHGCL framework relative to conventional baseline methods. Our model implementation is available at https://github.com/wangyu0627/IHGCL.
Despite recent advancements in Self-Supervised Learning (SSL) for time series analysis, a noticeable gap persists between the anticipated achievements and actual performance. While these methods have demonstrated formidable generalization capabilities with minimal labels in various domains, their effectiveness in distinguishing between different classes based on a limited number of annotated records is notably lacking. Our hypothesis attributes this bottleneck to the prevalent use of Contrastive Learning, a shared training objective in previous state-of-the-art (SOTA) methods. By mandating distinctiveness between representations for negative pairs drawn from separate records, this approach compels the model to encode unique record-based patterns but simultaneously neglects changes occurring across the entire record. To overcome this challenge, we introduce Distilled Embedding for Almost-Periodic Time Series (DEAPS) in this paper, offering a non-contrastive method tailored for quasiperiodic time series, such as electrocardiogram (ECG) data. By avoiding the use of negative pairs, we not only mitigate the model's blindness to temporal changes but also enable the integration of a "Gradual Loss (Lgra)" function. This function guides the model to effectively capture dynamic patterns evolving throughout the record. The outcomes are promising, as DEAPS demonstrates a notable improvement of +10% over existing SOTA methods when just a few annotated records are presented to fit a Machine Learning (ML) model based on the learned representation.