Mobile Edge Caching (MEC) integrated with Deep Neural Networks (DNNs) is an innovative technology with significant potential for the future generation of wireless networks, resulting in a considerable reduction in users' latency. The MEC network's effectiveness, however, heavily relies on its capacity to predict and dynamically update the storage of caching nodes with the most popular contents. To be effective, a DNN-based popularity prediction model needs to have the ability to understand the historical request patterns of content, including their temporal and spatial correlations. Existing state-of-the-art time-series DNN models capture the latter by simultaneously inputting the sequential request patterns of multiple contents to the network, considerably increasing the size of the input sample. This motivates us to address this challenge by proposing a DNN-based popularity prediction framework based on the idea of contrasting input samples against each other, designed for the Unmanned Aerial Vehicle (UAV)-aided MEC networks. Referred to as the Contrastive Learning-based Survival Analysis (CLSA), the proposed architecture consists of a self-supervised Contrastive Learning (CL) model, where the temporal information of sequential requests is learned using a Long Short Term Memory (LSTM) network as the encoder of the CL architecture. Followed by a Survival Analysis (SA) network, the output of the proposed CLSA architecture is probabilities for each content's future popularity, which are then sorted in descending order to identify the Top-K popular contents. Based on the simulation results, the proposed CLSA architecture outperforms its counterparts across the classification accuracy and cache-hit ratio.
Deep learning has achieved great success in recent years with the aid of advanced neural network structures and large-scale human-annotated datasets. However, it is often costly and difficult to accurately and efficiently annotate large-scale datasets, especially for some specialized domains where fine-grained labels are required. In this setting, coarse labels are much easier to acquire as they do not require expert knowledge. In this work, we propose a contrastive learning method, called $\textbf{Mask}$ed $\textbf{Con}$trastive learning~($\textbf{MaskCon}$) to address the under-explored problem setting, where we learn with a coarse-labelled dataset in order to address a finer labelling problem. More specifically, within the contrastive learning framework, for each sample our method generates soft-labels with the aid of coarse labels against other samples and another augmented view of the sample in question. By contrast to self-supervised contrastive learning where only the sample's augmentations are considered hard positives, and in supervised contrastive learning where only samples with the same coarse labels are considered hard positives, we propose soft labels based on sample distances, that are masked by the coarse labels. This allows us to utilize both inter-sample relations and coarse labels. We demonstrate that our method can obtain as special cases many existing state-of-the-art works and that it provides tighter bounds on the generalization error. Experimentally, our method achieves significant improvement over the current state-of-the-art in various datasets, including CIFAR10, CIFAR100, ImageNet-1K, Standford Online Products and Stanford Cars196 datasets. Code and annotations are available at https://github.com/MrChenFeng/MaskCon_CVPR2023.
Renal transplantation emerges as the most effective solution for end-stage renal disease. Occurring from complex causes, a substantial risk of transplant chronic dysfunction persists and may lead to graft loss. Medical imaging plays a substantial role in renal transplant monitoring in clinical practice. However, graft supervision is multi-disciplinary, notably joining nephrology, urology, and radiology, while identifying robust biomarkers from such high-dimensional and complex data for prognosis is challenging. In this work, taking inspiration from the recent success of Large Language Models (LLMs), we propose MEDIMP -- Medical Images with clinical Prompts -- a model to learn meaningful multi-modal representations of renal transplant Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE MRI) by incorporating structural clinicobiological data after translating them into text prompts. MEDIMP is based on contrastive learning from joint text-image paired embeddings to perform this challenging task. Moreover, we propose a framework that generates medical prompts using automatic textual data augmentations from LLMs. Our goal is to learn meaningful manifolds of renal transplant DCE MRI, interesting for the prognosis of the transplant or patient status (2, 3, and 4 years after the transplant), fully exploiting the limited available multi-modal data most efficiently. Extensive experiments and comparisons with other renal transplant representation learning methods with limited data prove the effectiveness of MEDIMP in a relevant clinical setting, giving new directions toward medical prompts. Our code is available at https://github.com/leomlck/MEDIMP.
Current multimodal sequential recommendation models are often unable to effectively explore and capture correlations among behavior sequences of users and items across different modalities, either neglecting correlations among sequence representations or inadequately capturing associations between multimodal data and sequence data in their representations. To address this problem, we explore multimodal pre-training in the context of sequential recommendation, with the aim of enhancing fusion and utilization of multimodal information. We propose a novel Multimodal Pre-training for Sequential Recommendation (MP4SR) framework, which utilizes contrastive losses to capture the correlation among different modality sequences of users, as well as the correlation among different modality sequences of users and items. MP4SR consists of three key components: 1) multimodal feature extraction, 2) a backbone network, Multimodal Mixup Sequence Encoder (M2SE), and 3) pre-training tasks. After utilizing pre-trained encoders to generate initial multimodal features of items, M2SE adopts a complementary sequence mixup strategy to fuse different modality sequences, and leverages contrastive learning to capture modality interactions at the sequence-to-sequence and sequence-to-item levels. Extensive experiments on four real-world datasets demonstrate that MP4SR outperforms state-of-the-art approaches in both normal and cold-start settings. We further highlight the efficacy of incorporating multimodal pre-training in sequential recommendation representation learning, serving as an effective regularizer and optimizing the parameter space for the recommendation task.
Current sequential recommender systems are proposed to tackle the dynamic user preference learning with various neural techniques, such as Transformer and Graph Neural Networks (GNNs). However, inference from the highly sparse user behavior data may hinder the representation ability of sequential pattern encoding. To address the label shortage issue, contrastive learning (CL) methods are proposed recently to perform data augmentation in two fashions: (i) randomly corrupting the sequence data (e.g. stochastic masking, reordering); (ii) aligning representations across pre-defined contrastive views. Although effective, we argue that current CL-based methods have limitations in addressing popularity bias and disentangling of user conformity and real interest. In this paper, we propose a new Debiased Contrastive learning paradigm for Recommendation (DCRec) that unifies sequential pattern encoding with global collaborative relation modeling through adaptive conformity-aware augmentation. This solution is designed to tackle the popularity bias issue in recommendation systems. Our debiased contrastive learning framework effectively captures both the patterns of item transitions within sequences and the dependencies between users across sequences. Our experiments on various real-world datasets have demonstrated that DCRec significantly outperforms state-of-the-art baselines, indicating its efficacy for recommendation. To facilitate reproducibility of our results, we make our implementation of DCRec publicly available at: https://github.com/HKUDS/DCRec.
Various contrastive learning approaches have been proposed in recent years and achieve significant empirical success. While effective and prevalent, contrastive learning has been less explored for time series data. A key component of contrastive learning is to select appropriate augmentations imposing some priors to construct feasible positive samples, such that an encoder can be trained to learn robust and discriminative representations. Unlike image and language domains where ``desired'' augmented samples can be generated with the rule of thumb guided by prefabricated human priors, the ad-hoc manual selection of time series augmentations is hindered by their diverse and human-unrecognizable temporal structures. How to find the desired augmentations of time series data that are meaningful for given contrastive learning tasks and datasets remains an open question. In this work, we address the problem by encouraging both high \textit{fidelity} and \textit{variety} based upon information theory. A theoretical analysis leads to the criteria for selecting feasible data augmentations. On top of that, we propose a new contrastive learning approach with information-aware augmentations, InfoTS, that adaptively selects optimal augmentations for time series representation learning. Experiments on various datasets show highly competitive performance with up to 12.0\% reduction in MSE on forecasting tasks and up to 3.7\% relative improvement in accuracy on classification tasks over the leading baselines.
Lack of audio-video synchronization is a common problem during television broadcasts and video conferencing, leading to an unsatisfactory viewing experience. A widely accepted paradigm is to create an error detection mechanism that identifies the cases when audio is leading or lagging. We propose ModEFormer, which independently extracts audio and video embeddings using modality-specific transformers. Different from the other transformer-based approaches, ModEFormer preserves the modality of the input streams which allows us to use a larger batch size with more negative audio samples for contrastive learning. Further, we propose a trade-off between the number of negative samples and number of unique samples in a batch to significantly exceed the performance of previous methods. Experimental results show that ModEFormer achieves state-of-the-art performance, 94.5% for LRS2 and 90.9% for LRS3. Finally, we demonstrate how ModEFormer can be used for offset detection for test clips.
We propose ViC-MAE, a model that combines both Masked AutoEncoders (MAE) and contrastive learning. ViC-MAE is trained using a global featured obtained by pooling the local representations learned under an MAE reconstruction loss and leveraging this representation under a contrastive objective across images and video frames. We show that visual representations learned under ViC-MAE generalize well to both video and image classification tasks. Particularly, ViC-MAE obtains state-of-the-art transfer learning performance from video to images on Imagenet-1k compared to the recently proposed OmniMAE by achieving a top-1 accuracy of 86% (+1.3% absolute improvement) when trained on the same data and 87.1% (+2.4% absolute improvement) when training on extra data. At the same time ViC-MAE outperforms most other methods on video benchmarks by obtaining 75.9% top-1 accuracy on the challenging Something something-v2 video benchmark . When training on videos and images from a diverse combination of datasets, our method maintains a balanced transfer-learning performance between video and image classification benchmarks, coming only as a close second to the best supervised method.
Massive rumors usually appear along with breaking news or trending topics, seriously hindering the truth. Existing rumor detection methods are mostly focused on the same domain, and thus have poor performance in cross-domain scenarios due to domain shift. In this work, we propose an end-to-end instance-wise and prototype-wise contrastive learning model with a cross-attention mechanism for cross-domain rumor detection. The model not only performs cross-domain feature alignment but also enforces target samples to align with the corresponding prototypes of a given source domain. Since target labels in a target domain are unavailable, we use a clustering-based approach with carefully initialized centers by a batch of source domain samples to produce pseudo labels. Moreover, we use a cross-attention mechanism on a pair of source data and target data with the same labels to learn domain-invariant representations. Because the samples in a domain pair tend to express similar semantic patterns, especially on the people's attitudes (e.g., supporting or denying) towards the same category of rumors, the discrepancy between a pair of the source domain and target domain will be decreased. We conduct experiments on four groups of cross-domain datasets and show that our proposed model achieves state-of-the-art performance.
We tackle the task of text-to-3D creation with pre-trained latent-based NeRFs (NeRFs that generate 3D objects given input latent code). Recent works such as DreamFusion and Magic3D have shown great success in generating 3D content using NeRFs and text prompts, but the current approach of optimizing a NeRF for every text prompt is 1) extremely time-consuming and 2) often leads to low-resolution outputs. To address these challenges, we propose a novel method named 3D-CLFusion which leverages the pre-trained latent-based NeRFs and performs fast 3D content creation in less than a minute. In particular, we introduce a latent diffusion prior network for learning the w latent from the input CLIP text/image embeddings. This pipeline allows us to produce the w latent without further optimization during inference and the pre-trained NeRF is able to perform multi-view high-resolution 3D synthesis based on the latent. We note that the novelty of our model lies in that we introduce contrastive learning during training the diffusion prior which enables the generation of the valid view-invariant latent code. We demonstrate through experiments the effectiveness of our proposed view-invariant diffusion process for fast text-to-3D creation, e.g., 100 times faster than DreamFusion. We note that our model is able to serve as the role of a plug-and-play tool for text-to-3D with pre-trained NeRFs.
Cross-View Geo-Localisation is still a challenging task where additional modules, specific pre-processing or zooming strategies are necessary to determine accurate positions of images. Since different views have different geometries, pre-processing like polar transformation helps to merge them. However, this results in distorted images which then have to be rectified. Adding hard negatives to the training batch could improve the overall performance but with the default loss functions in geo-localisation it is difficult to include them. In this article, we present a simplified but effective architecture based on contrastive learning with symmetric InfoNCE loss that outperforms current state-of-the-art results. Our framework consists of a narrow training pipeline that eliminates the need of using aggregation modules, avoids further pre-processing steps and even increases the generalisation capability of the model to unknown regions. We introduce two types of sampling strategies for hard negatives. The first explicitly exploits geographically neighboring locations to provide a good starting point. The second leverages the visual similarity between the image embeddings in order to mine hard negative samples. Our work shows excellent performance on common cross-view datasets like CVUSA, CVACT, University-1652 and VIGOR. A comparison between cross-area and same-area settings demonstrate the good generalisation capability of our model.
Spatial transcriptomic (ST) clustering employs spatial and transcription information to group spots spatially coherent and transcriptionally similar together into the same spatial domain. Graph convolution network (GCN) and graph attention network (GAT), fed with spatial coordinates derived adjacency and transcription profile derived feature matrix are often used to solve the problem. Our proposed method STGIC (spatial transcriptomic clustering with graph and image convolution) utilizes an adaptive graph convolution (AGC) to get high quality pseudo-labels and then resorts to dilated convolution framework (DCF) for virtual image converted from gene expression information and spatial coordinates of spots. The dilation rates and kernel sizes are set appropriately and updating of weight values in the kernels is made to be subject to the spatial distance from the position of corresponding elements to kernel centers so that feature extraction of each spot is better guided by spatial distance to neighbor spots. Self-supervision realized by KL-divergence, spatial continuity loss and cross entropy calculated among spots with high confidence pseudo-labels make up the training objective of DCF. STGIC attains state-of-the-art (SOTA) clustering performance on the benchmark dataset of human dorsolateral prefrontal cortex (DLPFC). Besides, it's capable of depicting fine structures of other tissues from other species as well as guiding the identification of marker genes. Also, STGIC is expandable to Stereo-seq data with high spatial resolution.
In this paper, we investigate representation learning for low-resource keyword spotting (KWS). The main challenges of KWS are limited labeled data and limited available device resources. To address those challenges, we explore representation learning for KWS by self-supervised contrastive learning and self-training with pretrained model. First, local-global contrastive siamese networks (LGCSiam) are designed to learn similar utterance-level representations for similar audio samplers by proposed local-global contrastive loss without requiring ground-truth. Second, a self-supervised pretrained Wav2Vec 2.0 model is applied as a constraint module (WVC) to force the KWS model to learn frame-level acoustic representations. By the LGCSiam and WVC modules, the proposed small-footprint KWS model can be pretrained with unlabeled data. Experiments on speech commands dataset show that the self-training WVC module and the self-supervised LGCSiam module significantly improve accuracy, especially in the case of training on a small labeled dataset.
Masked image modeling (MIM) with transformer backbones has recently been exploited as a powerful self-supervised pre-training technique. The existing MIM methods adopt the strategy to mask random patches of the image and reconstruct the missing pixels, which only considers semantic information at a lower level, and causes a long pre-training time.This paper presents HybridMIM, a novel hybrid self-supervised learning method based on masked image modeling for 3D medical image segmentation.Specifically, we design a two-level masking hierarchy to specify which and how patches in sub-volumes are masked, effectively providing the constraints of higher level semantic information. Then we learn the semantic information of medical images at three levels, including:1) partial region prediction to reconstruct key contents of the 3D image, which largely reduces the pre-training time burden (pixel-level); 2) patch-masking perception to learn the spatial relationship between the patches in each sub-volume (region-level).and 3) drop-out-based contrastive learning between samples within a mini-batch, which further improves the generalization ability of the framework (sample-level). The proposed framework is versatile to support both CNN and transformer as encoder backbones, and also enables to pre-train decoders for image segmentation. We conduct comprehensive experiments on four widely-used public medical image segmentation datasets, including BraTS2020, BTCV, MSD Liver, and MSD Spleen. The experimental results show the clear superiority of HybridMIM against competing supervised methods, masked pre-training approaches, and other self-supervised methods, in terms of quantitative metrics, timing performance and qualitative observations. The codes of HybridMIM are available at https://github.com/ge-xing/HybridMIM
Automatic radiology reporting has great clinical potential to relieve radiologists from heavy workloads and improve diagnosis interpretation. Recently, researchers have enhanced data-driven neural networks with medical knowledge graphs to eliminate the severe visual and textual bias in this task. The structures of such graphs are exploited by using the clinical dependencies formed by the disease topic tags via general knowledge and usually do not update during the training process. Consequently, the fixed graphs can not guarantee the most appropriate scope of knowledge and limit the effectiveness. To address the limitation, we propose a knowledge graph with Dynamic structure and nodes to facilitate medical report generation with Contrastive Learning, named DCL. In detail, the fundamental structure of our graph is pre-constructed from general knowledge. Then we explore specific knowledge extracted from the retrieved reports to add additional nodes or redefine their relations in a bottom-up manner. Each image feature is integrated with its very own updated graph before being fed into the decoder module for report generation. Finally, this paper introduces Image-Report Contrastive and Image-Report Matching losses to better represent visual features and textual information. Evaluated on IU-Xray and MIMIC-CXR datasets, our DCL outperforms previous state-of-the-art models on these two benchmarks.
Although many recent works have investigated generalizable NeRF-based novel view synthesis for unseen scenes, they seldom consider the synthetic-to-real generalization, which is desired in many practical applications. In this work, we first investigate the effects of synthetic data in synthetic-to-real novel view synthesis and surprisingly observe that models trained with synthetic data tend to produce sharper but less accurate volume densities. For pixels where the volume densities are correct, fine-grained details will be obtained. Otherwise, severe artifacts will be produced. To maintain the advantages of using synthetic data while avoiding its negative effects, we propose to introduce geometry-aware contrastive learning to learn multi-view consistent features with geometric constraints. Meanwhile, we adopt cross-view attention to further enhance the geometry perception of features by querying features across input views. Experiments demonstrate that under the synthetic-to-real setting, our method can render images with higher quality and better fine-grained details, outperforming existing generalizable novel view synthesis methods in terms of PSNR, SSIM, and LPIPS. When trained on real data, our method also achieves state-of-the-art results.
The self-supervised pretraining paradigm has achieved great success in skeleton-based action recognition. However, these methods treat the motion and static parts equally, and lack an adaptive design for different parts, which has a negative impact on the accuracy of action recognition. To realize the adaptive action modeling of both parts, we propose an Actionlet-Dependent Contrastive Learning method (ActCLR). The actionlet, defined as the discriminative subset of the human skeleton, effectively decomposes motion regions for better action modeling. In detail, by contrasting with the static anchor without motion, we extract the motion region of the skeleton data, which serves as the actionlet, in an unsupervised manner. Then, centering on actionlet, a motion-adaptive data transformation method is built. Different data transformations are applied to actionlet and non-actionlet regions to introduce more diversity while maintaining their own characteristics. Meanwhile, we propose a semantic-aware feature pooling method to build feature representations among motion and static regions in a distinguished manner. Extensive experiments on NTU RGB+D and PKUMMD show that the proposed method achieves remarkable action recognition performance. More visualization and quantitative experiments demonstrate the effectiveness of our method. Our project website is available at https://langlandslin.github.io/projects/ActCLR/
Text segmentation is a challenging vision task with many downstream applications. Current text segmentation methods require pixel-level annotations, which are expensive in the cost of human labor and limited in application scenarios. In this paper, we take the first attempt to perform weakly-supervised text instance segmentation by bridging text recognition and text segmentation. The insight is that text recognition methods provide precise attention position of each text instance, and the attention location can feed to both a text adaptive refinement head (TAR) and a text segmentation head. Specifically, the proposed TAR generates pseudo labels by performing two-stage iterative refinement operations on the attention location to fit the accurate boundaries of the corresponding text instance. Meanwhile, the text segmentation head takes the rough attention location to predict segmentation masks which are supervised by the aforementioned pseudo labels. In addition, we design a mask-augmented contrastive learning by treating our segmentation result as an augmented version of the input text image, thus improving the visual representation and further enhancing the performance of both recognition and segmentation. The experimental results demonstrate that the proposed method significantly outperforms weakly-supervised instance segmentation methods on ICDAR13-FST (18.95$\%$ improvement) and TextSeg (17.80$\%$ improvement) benchmarks.
Multifold observations are common for different data modalities, e.g., a 3D shape can be represented by multi-view images and an image can be described with different captions. Existing cross-modal contrastive representation learning (XM-CLR) methods such as CLIP are not fully suitable for multifold data as they only consider one positive pair and treat other pairs as negative when computing the contrastive loss. In this paper, we propose MXM-CLR, a unified framework for contrastive learning of multifold cross-modal representations. MXM-CLR explicitly models and learns the relationships between multifold observations of instances from different modalities for more comprehensive representation learning. The key of MXM-CLR is a novel multifold-aware hybrid loss which considers multiple positive observations when computing the hard and soft relationships for the cross-modal data pairs. We conduct quantitative and qualitative comparisons with SOTA baselines for cross-modal retrieval tasks on the Text2Shape and Flickr30K datasets. We also perform extensive evaluations on the adaptability and generalizability of MXM-CLR, as well as ablation studies on the loss design and effects of batch sizes. The results show the superiority of MXM-CLR in learning better representations for the multifold data. The code is available at https://github.com/JLU-ICL/MXM-CLR.
Previous gait recognition methods primarily trained on labeled datasets, which require painful labeling effort. However, using a pre-trained model on a new dataset without fine-tuning can lead to significant performance degradation. So to make the pre-trained gait recognition model able to be fine-tuned on unlabeled datasets, we propose a new task: Unsupervised Gait Recognition (UGR). We introduce a new cluster-based baseline to solve UGR with cluster-level contrastive learning. But we further find more challenges this task meets. First, sequences of the same person in different clothes tend to cluster separately due to the significant appearance changes. Second, sequences taken from 0{\deg} and 180{\deg} views lack walking postures and do not cluster with sequences taken from other views. To address these challenges, we propose a Selective Fusion method, which includes Selective Cluster Fusion (SCF) and Selective Sample Fusion (SSF). With SCF, we merge matched clusters of the same person wearing different clothes by updating the cluster-level memory bank with a multi-cluster update strategy. And in SSF, we merge sequences taken from front/back views gradually with curriculum learning. Extensive experiments show the effectiveness of our method in improving the rank-1 accuracy in walking with different coats condition and front/back views conditions.