Point cloud analysis is challenging due to its unique characteristics of unorderness, sparsity and irregularity. Prior works attempt to capture local relationships by convolution operations or attention mechanisms, exploiting geometric information from coordinates implicitly. These methods, however, are insufficient to describe the explicit local geometry, e.g., curvature and orientation. In this paper, we propose On-the-fly Point Feature Representation (OPFR), which captures abundant geometric information explicitly through Curve Feature Generator module. This is inspired by Point Feature Histogram (PFH) from computer vision community. However, the utilization of vanilla PFH encounters great difficulties when applied to large datasets and dense point clouds, as it demands considerable time for feature generation. In contrast, we introduce the Local Reference Constructor module, which approximates the local coordinate systems based on triangle sets. Owing to this, our OPFR only requires extra 1.56ms for inference (65X faster than vanilla PFH) and 0.012M more parameters, and it can serve as a versatile plug-and-play module for various backbones, particularly MLP-based and Transformer-based backbones examined in this study. Additionally, we introduce the novel Hierarchical Sampling module aimed at enhancing the quality of triangle sets, thereby ensuring robustness of the obtained geometric features. Our proposed method improves overall accuracy (OA) on ModelNet40 from 90.7% to 94.5% (+3.8%) for classification, and OA on S3DIS Area-5 from 86.4% to 90.0% (+3.6%) for semantic segmentation, respectively, building upon PointNet++ backbone. When integrated with Point Transformer backbone, we achieve state-of-the-art results on both tasks: 94.8% OA on ModelNet40 and 91.7% OA on S3DIS Area-5.
论文检索
输入标题、作者或关键词，从 7,876 篇学术成果中精准定位
Few-shot fine-grained image classification aims to use only few labelled samples to successfully recognize subtle sub-classes within the same parent class. This task is extremely challenging, due to the co-occurrence of large inter-class similarity, low intra-class similarity, and only few labelled samples. In this paper, to address these challenges, we propose a new Channel-Spatial Cross-Attention Module (CSCAM), which can effectively drive a model to extract discriminative fine-grained feature representations with only few shots. CSCAM collaboratively integrates a channel cross-attention module and a spatial cross-attention module, for the attentions across support and query samples. In addition, to fit for the characteristics of fine-grained images, a support averaging method is proposed in CSCAM to reduce the intra-class distance and increase the inter-class distance. Extensive experiments on four few-shot fine-grained classification datasets validate the effectiveness of CSCAM. Furthermore, CSCAM is a plug-and-play module, conveniently enabling effective improvement of state-of-the-art methods for few-shot fine-grained image classification.
Reducing the atmospheric haze and enhancing image clarity is crucial for computer vision applications. The lack of real-life hazy ground truth images necessitates synthetic datasets, which often lack diverse haze types, impeding effective haze type classification and dehazing algorithm selection. This research introduces the HazeSpace2M dataset, a collection of over 2 million images designed to enhance dehazing through haze type classification. HazeSpace2M includes diverse scenes with 10 haze intensity levels, featuring Fog, Cloud, and Environmental Haze (EH). Using the dataset, we introduce a technique of haze type classification followed by specialized dehazers to clear hazy images. Unlike conventional methods, our approach classifies haze types before applying type-specific dehazing, improving clarity in real-life hazy images. Benchmarking with state-of-the-art (SOTA) models, ResNet50 and AlexNet achieve 92.75% and 92.50% accuracy, respectively, against existing synthetic datasets. However, these models achieve only 80% and 70% accuracy, respectively, against our Real Hazy Testset (RHT), highlighting the challenging nature of our HazeSpace2M dataset. Additional experiments show that haze type classification followed by specialized dehazing improves results by 2.41% in PSNR, 17.14% in SSIM, and 10.2% in MSE over general dehazers. Moreover, when testing with SOTA dehazing models, we found that applying our proposed framework significantly improves their performance. These results underscore the significance of HazeSpace2M and our proposed framework in addressing atmospheric haze in multimedia processing. Complete code and dataset is available on GitHub (https://github.com/tanvirnwu/HazeSpace2M).
Most existing NAS-based multi-modal classification (MMC-NAS) methods are optimized using the classification accuracy.They can not simultaneously provide multiple models with diverse perferences such as model complex and classification performance for meeting different users' demands. Combining NAS-MMC with multi-objective optimization is a nature way for this issue. However, the challenge problem of this solution is the high computation cost. For multi-objective optimization, the computing bottleneck is pareto front search. Some higher-quality MMC models (namely core structures, CSs) consisting of high-quality features and fusion operators are easier to identify. We find that CSs have a close relation with the pareto front (PF), i.e., the individuals lying in PF contain the CSs. Based on the finding, we propose an efficient multi-objective neural architecture search for multi-modal classification by applying CSs to guide the PF search (CoMO-NAS). In conclusion, experimental results thoroughly demonstrate the effectiveness of our CoMO-NAS. Compared to state-of-the-art competitors on benchmark multi-modal tasks, we achieve comparable performance with lower model complexity in shorter search time.
In the realm of CLIP adaptation through prompt learning, it is important to emphasize the pivotal role that the proper alignment of visual and textual representations plays when adapting the CLIP to downstream tasks. We propose that the proper alignment for downstream tasks is determined by the flexibility of the interaction between cross-modal information, which compensates for the absence of contrastive loss during the adaptation process. However, the current prompt learning methods, such as isolated modifications to the visual or language branches of CLIP or the employment of uni-directional cross-modal fusion, are not sufficient to explore the full potential of the mutual interaction between visual and textual modalities. To overcome this limitation, we propose a new paradigm for the CLIP prompt learning community, named Bilateral Adaptive Cross-Modal Fusion Prompt Learning (Bloom), which includes two enhancements. First, we propose using projection functions for bi-directional modality transformation and fusion functions to encourage the mutual interaction between corresponding layers within both the image and text encoders. Second, we propose an adaptive manner that automatically searches the optimal combination of cross-modal information at each layer. These two improvements ensure a more efficient and flexible integration of the two modalities, thereby achieving proper alignment for specific downstream tasks. We put our method to the test in terms of base-to-novel, cross-dataset, and cross-domain evaluations on 15 image classification datasets. The results demonstrate a significant performance enhancement achieved by Bloom.
Decoding human visual representations from brain activity data is a challenging but arguably essential task with an understanding of the real world and the human visual system. However, decoding semantically similar visual representations from brain recordings is difficult, especially for electroencephalography (EEG), which has excellent temporal resolution but suffers from spatial precision. Prevailing methods mainly focus on matching brain activity data with corresponding stimuli-responses using contrastive learning. They rely on massive and high-quality paired data and omit semantically aligned modalities distributed in distinct regions of the latent space. This paper proposes a novel Multimodal Bidirectional Cycle Consistency (MB2C) framework for learning robust visual neural representations. Specifically, we utilize dual-GAN to generate modality-related features and inversely translate back to the corresponding semantic latent space to close the modality gap and guarantee that embeddings from different modalities with similar semantics are in the same region of representation space. We perform zero-shot tasks on the ThingsEEG dataset. Additionally, we conduct EEG classification and image reconstruction on both the ThingsEEG and EEGCVPR40 datasets, achieving state-of-the-art performance compared to other baselines.
Continual learning emerges as a framework that trains the model on a sequence of tasks without forgetting previously learned knowledge, which has been applied in multiple multimodal scenarios. Recently, prompt-based continual learning has achieved excellent domain adaptability and knowledge transfer through prompt generation. However, existing methods mainly focus on designing the architecture of a generator, neglecting the importance of providing effective guidance for training the generator. To address this issue, we propose Generating Prompts in Latent Space (GPLS), which considers prompts as latent variables to account for the uncertainty of prompt generation and aligns with the fact that prompts are inserted into the hidden layer outputs and exert an implicit influence on classification. GPLS adopts a trainable encoder to encode task and feature information into prompts with reparameterization technique, and provides refined and targeted guidance for the training process through the evidence lower bound (ELBO) related to Mahalanobis distance. Extensive experiments demonstrate that GPLS achieves state-of-the-art performance on various benchmarks. Our code is available at https://github.com/Hifipsysta/GPLS.
Image-to-Video adaptation is proposed to train a model using labeled images and unlabeled videos to facilitate the classification of unlabeled videos. The latest work synthesizes videos using still images to mitigate the modality gap between images and videos. However, the synthesized videos are not realistic due to the camera movements are only simulated in 2D space. Therefore, we generate realistic videos by simulating arbitrary camera movements in 3D scenes, and then the model can be trained using the generated source videos. Unfortunately, the optical flows from the generated videos have unexpected negative impacts, resulting in suboptimal performance. To address this issue, we propose the Category-aware Flow Memory Bank, which replaces optical flows in source videos with real target flows, and the new composed videos are beneficial for training. In addition, we leverage the video pace prediction task to enhance the model's perception of speed. Our method achieves state-of-the-art performance and comparable performance on three widely used benchmarks. Our code is available at https://github.com/KenanHuang/mm2024\_cfb4i2v.
In recent years, Few-Shot Object Detection (FSOD) has gained widespread attention and made significant progress due to its ability to build models with a good generalization power using extremely limited annotated data. The fine-tuning based paradigm is currently dominating this field, where detectors are initially pre-trained on base classes with sufficient samples and then fine-tuned on novel ones with few samples, but the scarcity of labeled samples of novel classes greatly interferes precisely fitting their data distribution, thus hampering the performance. To address this issue, we propose a new framework for FSOD, namely Prototype-based Soft-labels and Test-Time Learning (PS-TTL). Specifically, we design a Test-Time Learning (TTL) module that employs a mean-teacher network for self-training to discover novel instances from test data, allowing detectors to learn better representations and classifiers for novel classes. Furthermore, we notice that even though relatively low-confidence pseudo-labels exhibit classification confusion, they still tend to recall foreground. We thus develop a Prototype-based Soft-labels (PS) strategy through assessing similarities between low-confidence pseudo-labels and category prototypes as soft-labels to unleash their potential, which substantially mitigates the constraints posed by few-shot samples. Extensive experiments on both the VOC and COCO benchmarks show that PS-TTL achieves the state-of-the-art, highlighting its effectiveness. The code and model are available at https://github.com/gaoyingjay/PS-TTL.
Handling varying computational resources is a critical issue in modern AI applications. Adaptive deep networks, featuring the dynamic employment of multiple classifier heads among different layers, have been proposed to address classification tasks under varying computing resources. Existing approaches typically utilize the last classifier supported by the available resources for inference, as they believe that the last classifier always performs better across all classes. However, our findings indicate that earlier classifier heads can outperform the last head for certain classes. Based on this observation, we introduce the Collaborative Decision Making (CDM) module, which fuses the multiple classifier heads to enhance the inference performance of adaptive deep networks. CDM incorporates an uncertainty-aware fusion method based on evidential deep learning (EDL), that utilizes the reliability (uncertainty values) from the first c-1 classifiers to improve the c-th classifier' accuracy. We also design a balance term that reduces fusion saturation and unfairness issues caused by EDL constraints to improve the fusion quality of CDM. Finally, a regularized training strategy that uses the last classifier to guide the learning process of early classifiers is proposed to further enhance the CDM module's effect, called the Guided Collaborative Decision Making (GCDM) framework. The experimental evaluation demonstrates the effectiveness of our approaches. Results on ImageNet datasets show CDM and GCDM obtain 0.4% to 2.8% accuracy improvement (under varying computing resources) on popular adaptive networks. The code is available at the link https://github.com/Meteor-Stars/GCDM_AdaptiveNet.
Automated Machine Learning (AutoML) offers a promising approach to streamline the training of machine learning models. However, existing AutoML frameworks are often limited to unimodal scenarios and require extensive manual configuration. Recent advancements in Large Language Models (LLMs) have showcased their exceptional abilities in reasoning, interaction, and code generation, presenting an opportunity to develop a more automated and user-friendly framework. To this end, we introduce AutoM3L, an innovative Automated Multimodal Machine Learning framework that leverages LLMs as controllers to automatically construct multimodal training pipelines. AutoM3L comprehends data modalities and selects appropriate models based on user requirements, providing automation and interactivity. By eliminating the need for manual feature engineering and hyperparameter optimization, our framework simplifies user engagement and enables customization through directives, addressing the limitations of previous rule-based AutoML approaches. We evaluate the performance of AutoM3L on six diverse multimodal datasets spanning classification, regression, and retrieval tasks, as well as a comprehensive set of unimodal datasets. The results demonstrate that AutoM3L achieves competitive or superior performance compared to traditional rule-based AutoML methods. Furthermore, a user study highlights the user-friendliness and usability of our framework, compared to the rule-based AutoML methods.
Multi-label image classification is crucial for a wide range of multimedia applications. To address the resource limitation issue, various knowledge distillation (KD) methods have been developed to transfer knowledge from a large network (referred to as the "teacher") to a small network (referred to as the "student"). However, existing KD methods do not explicitly distill the dependencies between labels, which limits the model ability to capture multi-label correlation. Furthermore, although existing methods for multi-label image classification have utilized the second-order label pair dependency (direct dependency between two labels), the high-order label pair dependency, which captures the indirect dependency between two labels, remains unexplored. In this paper, we propose a Multi-Order Label Pair Dependencies Knowledge Distillation (MDKD) framework. MDKD explicitly distills the knowledge to capture multi-order dependencies between labels, including the label pair dependencies from second-order and high-order, thus transferring the insight of label correlations from different perspectives. Extensive experiments on Pascal VOC2007, MSCOCO2014, and NUS-WIDE demonstrate the superior performances of MDKD.
In recent years, the Few-Shot Fine-Grained Image Classification (FS-FGIC) problem has gained widespread attention. A number of effective methods have been proposed that focus on extracting discriminative information within high-level features in a single episode/task. However, this is insufficient for addressing the cross-task challenges of FS-FGIC, which is represented in two aspects. On the one hand, from the perspective of the Fine-Grained Image Classification (FGIC) task, there is a need to supplement the model with mid-level features containing rich fine-grained information. On the other hand, from the perspective of the Few-Shot Learning (FSL) task, explicit modeling of cross-task general knowledge is required. In this paper, we propose a novel Bi-directional Task-Guided Network (BTG-Net) to tackle these issues. Specifically, from the FGIC task perspective, we design the Semantic-Guided Noise Filtering (SGNF) module to filter noise on mid-level features rich in detailed information. Further, from the FSL task perspective, the General Knowledge Prompt Modeling (GKPM) module is proposed to retain the cross-task general knowledge by utilizing the prompting mechanism, thereby enhancing the model's generalization performance on novel classes. We have conducted extensive experiments on five fine-grained benchmark datasets, and the results demonstrate that BTG-Net outperforms state-of-the-art methods comprehensively.
Weakly-supervised Temporal Action Localization (WTAL) following a localization-by-classification paradigm has achieved significant results, yet still grapples with confounding arising from ambiguous snippets. Previous works have attempted to distinguish these ambiguous snippets from action snippets without investigating the underlying causes of their formation, thus failing to effectively eliminate the bias on both action-context and action-content. In this paper, we revisit WTAL from the perspective of structural causal model to identify the true origins of confounding, and propose an efficient dual-confounding eliminating framework to alleviate these biases. Specifically, we construct a Substituted Confounder Set (SCS) to eliminate the confounding bias on action-context by leveraging the modal disparity between RGB and FLOW. Then, a Multi-level Consistency Mining (MCM) method is designed to mitigate the confounding bias on action-content by utilizing the consistency between discriminative snippets and corresponding proposals at both the feature and label levels. Notably, SCS and MCM could be seamlessly integrated into any two-stream models without additional parameters by Expectation-Maximization (EM) algorithm. Extensive experiments on two challenging benchmarks including THUMOS14 and ActivityNet-1.2 demonstrate the superior performance of our method.
Binarized Vision Transformers (BiViTs) aim to facilitate the efficient and lightweight utilization of Vision Transformers (ViTs) on devices with limited computational resources. Yet, the current approach to binarizing ViT leads to a substantial performance decrease compared to the full-precision model, posing obstacles to practical deployment. By empirical study, we reveal that spatial interaction (SI) is a critical factor that impacts performance due to lack of token-level correlation, but previous work ignores this factor. To this end, we design a ViT binarization approach dubbed SI-BiViT to incorporate spatial interaction in the binarization process. Specifically, an SI module is placed alongside the Multi-Layer Perceptron (MLP) module to formulate the dual-branch structure. This structure not only leverages knowledge from pre-trained ViTs by distilling over the original MLP, but also enhances spatial interaction via the introduced SI module. Correspondingly, we design a decoupled training strategy to train these two branches more effectively. Importantly, our SI-BiViT is orthogonal to existing Binarized ViTs approaches and can be directly plugged. Extensive experiments demonstrate the strong flexibility and effectiveness of SI-BiViT by plugging our method into four classic ViT backbones in supporting three downstream tasks, including classification, detection, and segmentation. In particular, SI-BiViT enhances the classification performance of binarized ViTs by an average of 10.52% in Top-1 accuracy compared to the previous state-of-the-art. Codes are available at https://github.com/VL-Group/SI-BiViT
Despite the advancements that deep learning has brought to medical image analysis (MIA), protecting the privacy of images remains a challenge. In a client-server MIA framework, especially after deployment, patients' private medical images can be easily captured by attackers from the transmission channel or malicious third-party servers. Previous MIA privacy-enhancing methods, whether based on distortion or homomorphic encryption, expose the fact that the transmitted images are medical images or transform the images into semantic-lacking noise. This tends to alert attackers, thereby falling into a cat-and-mouse game of theft and protection. To address this issue, we propose a covert MIA framework based on deep image hiding, namely HideMIA, which secures medical images by embedding them within natural cover images that are unlikely to raise suspicion. By directly analyzing the hidden medical images in the steganographic domain, HideMIA makes it difficult for attackers to notice the presence of medical images. Specifically, we propose the Mixture-of-Difference-Convolutions (MoDC) and Asymmetric Wavelet Attention (AsyWA) to enable HideMIA to conduct fine-grained analysis on each wavelet sub-band within the steganographic domain, mining features that are specific to medical images. Moreover, to reduce resource consumption on client devices, we design function-aligned knowledge distillation to obtain a lightweight hiding network, namely LightIH. Extensive experiments on six medical datasets demonstrate that our HideMIA achieves superior MIA performance and protective imperceptibility on medical image segmentation and classification.
The integration of large language models into open-world detection frameworks significantly improves versatility in new environments. Prompt representations derived from these models help establish classification boundaries for both base and novel categories within open-world detectors. However, we are the first to discover that directly fine-tuning language models in detection systems results in redundant attention patterns and leads to suboptimal prompt representations. In order to fully leverage the capabilities of large language models and augment prompt encoding for detection, this study introduces a redundancy assessment metric to identify uniform attention patterns. Furthermore, in areas with high redundancy, we incorporate multimodal inplace prompt tuning (MIPT) to enrich the text prompt with visual clues. Experimental results validate the efficacy of our MIPT framework, achieving a notable increase across benchmarks, e.g. elevating GLIP-L from 22.6% to 25.0% on ODinW-35, and 9.0% improvement on LVIS.
The wide use of mobile devices has led to a proliferated creation of extensive trajectory data, rendering trajectory classification increasingly vital and challenging for downstream applications. Existing deep learning methods offer powerful feature extraction capabilities to detect nuanced variances in trajectory classification tasks. However, their effectiveness remains compromised by the following two unsolved challenges. First, identifying the distribution of nearby trajectories based on noisy and sparse GPS coordinates poses a significant challenge, providing critical contextual features to the classification. Second, though efforts have been made to incorporate a shape feature by rendering trajectories into images, they fail to model the local correspondence between GPS points and image pixels. To address these issues, we propose a novel model termed Traj2Former to spotlight the spatial distribution of the adjacent trajectory points (i.e., contextual snapshot) and enhance the snapshot fusion between the trajectory data and the corresponding spatial contexts. We propose a new GPS rendering method to generate contextual snapshots, but it can be applied from a trajectory database to a digital map. Moreover, to capture diverse temporal patterns, we conduct a multi-scale sequential fusion by compressing the trajectory data with differing rates. Extensive experiments have been conducted to verify the superiority of the Traj2Former model.
Multimodal contrastive learning (MCL) has shown remarkable advances in zero-shot classification by learning from millions of image-caption pairs crawled from the Internet. However, this reliance poses privacy risks, as hackers may unauthorizedly exploit image-text data for model training, potentially including personal and privacy-sensitive information. Recent works propose generating unlearnable examples by adding imperceptible perturbations to training images to build shortcuts for protection. However, they are designed for unimodal classification, which remains largely unexplored in MCL. We first explore this context by evaluating the performance of existing methods on image-caption pairs, and they do not generalize effectively to multimodal data and exhibit limited impact to build shortcuts due to the lack of labels and the dispersion of pairs in MCL. In this paper, we propose Multi-step Error Minimization (MEM), a novel optimization process for generating multimodal unlearnable examples. It extends the Error-Minimization (EM) framework to optimize both image noise and an additional text trigger, thereby enlarging the optimized space and effectively misleading the model to learn the shortcut between the noise features and the text trigger. Specifically, we adopt projected gradient descent to solve the noise minimization problem and use HotFlip to approximate the gradient and replace words to find the optimal text trigger. Extensive experiments demonstrate the effectiveness of MEM, with post-protection retrieval results nearly half of random guessing, and its high transferability across different models. Our code is available on the https://github.com/thinwayliu/Multimodal-Unlearnable-Examples
Point clouds represent one of the prevalent formats for 3D content. Distortions introduced at various stages in the point cloud processing pipeline affect the visual quality, altering their geometric composition, texture information, or both. Understanding and quantifying the impact of the distortion domain on visual quality is vital to driving rate optimization and guiding post-processing steps to improve the quality of experience. In this paper, we propose a multi-task guided multi-modality no reference metric (M3-Unity), which utilizes 4 types of modalities across attributes and dimensionalities to represent point clouds. An attention mechanism establishes inter/intra associations among 3D/2D patches, which can complement each other, yielding local and global features, to fit the highly nonlinear property of the human vision system. A multi-task decoder involving distortion type classification selects the best association among 4 modalities, aiding the regression task and enabling the in-depth analysis of the interplay between geometrical and textural distortions. Furthermore, our framework design and attention strategy enable us to measure the impact of individual attributes and their combinations, providing insights into how these associations contribute particularly in relation to distortion type. Extensive experimental results on 4 datasets consistently outperform the state-of-the-art metrics by a large margin. The code is available at https://github.com/cwi-dis/ACMMM2024-Oral.