BEV perception is of great importance in the field of autonomous driving, serving as the cornerstone of planning, controlling, and motion prediction. The quality of the BEV feature highly affects the performance of BEV perception. However, taking the noises in camera parameters and LiDAR scans into consideration, we usually obtain BEV representation with harmful noises. Diffusion models naturally have the ability to denoise noisy samples to the ideal data, which motivates us to utilize the diffusion model to get a better BEV representation. In this work, we propose an end-to-end framework, named DiffBEV, to exploit the potential of diffusion model to generate a more comprehensive BEV representation. To the best of our knowledge, we are the first to apply diffusion model to BEV perception. In practice, we design three types of conditions to guide the training of the diffusion model which denoises the coarse samples and refines the semantic feature in a progressive way. What's more, a cross-attention module is leveraged to fuse the context of BEV feature and the semantic content of conditional diffusion model. DiffBEV achieves a 25.9% mIoU on the nuScenes dataset, which is 6.2% higher than the best-performing existing approach. Quantitative and qualitative results on multiple benchmarks demonstrate the effectiveness of DiffBEV in BEV semantic segmentation and 3D object detection tasks.
论文检索
输入标题、作者或关键词，从 11,272 篇学术成果中精准定位
IPRemover: A Generative Model Inversion Attack against Deep Neural Network Fingerprinting and Watermarking
PDF ↗Training Deep Neural Networks (DNNs) can be expensive when data is difficult to obtain or labeling them requires significant domain expertise. Hence, it is crucial that the Intellectual Property (IP) of DNNs trained on valuable data be protected against IP infringement. DNN fingerprinting and watermarking are two lines of work in DNN IP protection. Recently proposed DNN fingerprinting techniques are able to detect IP infringement while preserving model performance by relying on the key assumption that the decision boundaries of independently trained models are intrinsically different from one another. In contrast, DNN watermarking embeds a watermark in a model and verifies IP infringement if an identical or similar watermark is extracted from a suspect model. The techniques deployed in fingerprinting and watermarking vary significantly because their underlying mechanisms are different. From an adversary's perspective, a successful IP removal attack should defeat both fingerprinting and watermarking. However, to the best of our knowledge, there is no work on such attacks in the literature yet. In this paper, we fill this gap by presenting an IP removal attack that can defeat both fingerprinting and watermarking. We consider the challenging data-free scenario whereby all data is inverted from the victim model. Under this setting, a stolen model only depends on the victim model. Experimental results demonstrate the success of our attack in defeating state-of-the-art DNN fingerprinting and watermarking techniques. This work reveals a novel attack surface that exploits generative model inversion attacks to bypass DNN IP defenses. This threat must be addressed by future defenses for reliable IP protection.
Recent advances in implicit neural representation have demonstrated the ability to recover detailed geometry and material from multi-view images. However, the use of simplified lighting models such as environment maps to represent non-distant illumination, or using a network to fit indirect light modeling without a solid basis, can lead to an undesirable decomposition between lighting and material. To address this, we propose a fully differentiable framework named Neural Illumination Fields (NeIF) that uses radiance fields as a lighting model to handle complex lighting in a physically based way. Together with integral lobe encoding for roughness-adaptive specular lobe and leveraging the pre-convolved background for accurate decomposition, the proposed method represents a significant step towards integrating physically based rendering into the NeRF representation. The experiments demonstrate the superior performance of novel-view rendering compared to previous works, and the capability to re-render objects under arbitrary NeRF-style environments opens up exciting possibilities for bridging the gap between virtual and real-world scenes.
For the task of unsupervised image translation, transforming the image style while preserving its original structure remains challenging. In this paper, we propose an unsupervised image translation method with structural enhancement in frequency domain named SEIT. Specifically, a frequency dynamic adaptive (FDA) module is designed for image style transformation that can well transfer the image style while maintaining its overall structure by decoupling the image content and style in frequency domain. Moreover, a wavelet-based structure enhancement (WSE) module is proposed to improve the intermediate translation results by matching the high-frequency information, thus enriching the structural details. Furthermore, a multi-scale network architecture is designed to extract the domain-specific information using image-independent encoders for both the source and target domains. The extensive experimental results well demonstrate the effectiveness of the proposed method.
Current 3D object detection methods for indoor scenes mainly follow the voting-and-grouping strategy to generate proposals. However, most methods utilize instance-agnostic groupings, such as ball query, leading to inconsistent semantic information and inaccurate regression of the proposals. To this end, we propose a novel superpoint grouping network for indoor anchor-free one-stage 3D object detection. Specifically, we first adopt an unsupervised manner to partition raw point clouds into superpoints, areas with semantic consistency and spatial similarity. Then, we design a geometry-aware voting module that adapts to the centerness in anchor-free detection by constraining the spatial relationship between superpoints and object centers. Next, we present a superpoint-based grouping module to explore the consistent representation within proposals. This module includes a superpoint attention layer to learn feature interaction between neighboring superpoints, and a superpoint-voxel fusion layer to propagate the superpoint-level information to the voxel level. Finally, we employ effective multiple matching to capitalize on the dynamic receptive fields of proposals based on superpoints during the training. Experimental results demonstrate our method achieves state-of-the-art performance on ScanNet V2, SUN RGB-D, and S3DIS datasets in the indoor one-stage 3D object detection. Source code is available at https://github.com/zyrant/SPGroup3D.
Recent studies have made remarkable progress in histopathology classification. Based on current successes, contemporary works proposed to further upgrade the model towards a more generalizable and robust direction through incrementally learning from the sequentially delivered domains. Unlike previous parameter isolation based approaches that usually demand massive computation resources during model updating, we present a memory-efficient prompt tuning framework to cultivate model generalization potential in economical memory cost. For each incoming domain, we reuse the existing parameters of the initial classification model and attach lightweight trainable prompts into it for customized tuning. Considering the domain heterogeneity, we perform decoupled prompt tuning, where we adopt a domain-specific prompt for each domain to independently investigate its distinctive characteristics, and one domain-invariant prompt shared across all domains to continually explore the common content embedding throughout time. All domain-specific prompts will be appended to the prompt bank and isolated from further changes to prevent forgetting the distinctive features of early-seen domains. While the domain-invariant prompt will be passed on and iteratively evolve by style-augmented prompt refining to improve model generalization capability over time. In specific, we construct a graph with existing prompts and build a style-augmented graph attention network to guide the domain-invariant prompt exploring the overlapped latent embedding among all delivered domains for more domain-generic representations. We have extensively evaluated our framework with two histopathology tasks, i.e., breast cancer metastasis classification and epithelium-stroma tissue classification, where our approach yielded superior performance and memory efficiency over the competing methods.
Few-shot learning (FSL) based on manifold regularization aims to improve the recognition capacity of novel objects with limited training samples by mixing two samples from different categories with a blending factor. However, this mixing operation weakens the feature representation due to the linear interpolation and the overlooking of the importance of specific channels. To solve these issues, this paper proposes attentive feature regularization (AFR) which aims to improve the feature representativeness and discriminability. In our approach, we first calculate the relations between different categories of semantic labels to pick out the related features used for regularization. Then, we design two attention-based calculations at both the instance and channel levels. These calculations enable the regularization procedure to focus on two crucial aspects: the feature complementarity through adaptive interpolation in related categories and the emphasis on specific feature channels. Finally, we combine these regularization strategies to significantly improve the classifier performance. Empirical studies on several popular FSL benchmarks demonstrate the effectiveness of AFR, which improves the recognition accuracy of novel categories without the need to retrain any feature extractor, especially in the 1-shot setting. Furthermore, the proposed AFR can seamlessly integrate into other FSL methods to improve classification performance.
The goal of 3D mesh watermarking is to embed the message in 3D meshes that can withstand various attacks imperceptibly and reconstruct the message accurately from watermarked meshes. The watermarking algorithm is supposed to withstand multiple attacks, and the complexity should not grow significantly with the mesh size. Unfortunately, previous methods are less robust against attacks and lack of adaptability. In this paper, we propose a robust and adaptable deep 3D mesh watermarking Deep3DMark that leverages attention-based convolutions in watermarking tasks to embed binary messages in vertex distributions without texture assistance. Furthermore, our Deep3DMark exploits the property that simplified meshes inherit similar relations from the original ones, where the relation is the offset vector directed from one vertex to its neighbor. By doing so, our method can be trained on simplified meshes but remains effective on large size meshes (size adaptable) and unseen categories of meshes (geometry adaptable). Extensive experiments demonstrate our method remains efficient and effective even if the mesh size is 190× increased. Under mesh attacks, Deep3DMark achieves 10%∼50% higher accuracy than traditional methods, and 2× higher SNR and 8% higher accuracy than previous DNN-based methods.
Real-world text can be damaged by corrosion issues caused by environmental or human factors, which hinder the preservation of the complete styles of texts, e.g., texture and structure. These corrosion issues, such as graffiti signs and incomplete signatures, bring difficulties in understanding the texts, thereby posing significant challenges to downstream applications, e.g., scene text recognition and signature identification. Notably, current inpainting techniques often fail to adequately address this problem and have difficulties restoring accurate text images along with reasonable and consistent styles. Formulating this as an open problem of text image inpainting, this paper aims to build a benchmark to facilitate its study. In doing so, we establish two specific text inpainting datasets which contain scene text images and handwritten text images, respectively. Each of them includes images revamped by real-life and synthetic datasets, featuring pairs of original images, corrupted images, and other assistant information. On top of the datasets, we further develop a novel neural framework, Global Structure-guided Diffusion Model (GSDM), as a potential solution. Leveraging the global structure of the text as a prior, the proposed GSDM develops an efficient diffusion model to recover clean texts. The efficacy of our approach is demonstrated by thorough empirical study, including a substantial boost in both recognition accuracy and image quality. These findings not only highlight the effectiveness of our method but also underscore its potential to enhance the broader field of text image understanding and processing. Code and datasets are available at: https://github.com/blackprotoss/GSDM.
SEER: Backdoor Detection for Vision-Language Models through Searching Target Text and Image Trigger Jointly
PDF ↗This paper proposes SEER, a novel backdoor detection algorithm for vision-language models, addressing the gap in the literature on multi-modal backdoor detection. While backdoor detection in single-modal models has been well studied, the investigation of such defenses in multi-modal models remains limited. Existing backdoor defense mechanisms cannot be directly applied to multi-modal settings due to their increased complexity and search space explosion. In this paper, we propose to detect backdoors in vision-language models by jointly searching image triggers and malicious target texts in feature space shared by vision and language modalities. Our extensive experiments demonstrate that SEER can achieve over 92% detection rate on backdoor detection in vision-language models in various settings without accessing training data or knowledge of downstream tasks.
As a bio-inspired vision sensor, the spike camera emulates the operational principles of the fovea, a compact retinal region, by employing spike discharges to encode the accumulation of per-pixel luminance intensity. Leveraging its high temporal resolution and bio-inspired neuromorphic design, the spike camera holds significant promise for advancing computer vision applications. Saliency detection mimic the behavior of human beings and capture the most salient region from the scenes. In this paper, we investigate the visual saliency in the continuous spike stream for the first time. To effectively process the binary spike stream, we propose a Recurrent Spiking Transformer (RST) framework, which is based on a full spiking neural network. Our framework enables the extraction of spatio-temporal features from the continuous spatio-temporal spike stream while maintaining low power consumption. To facilitate the training and validation of our proposed model, we build a comprehensive real-world spike-based visual saliency dataset, enriched with numerous light conditions. Extensive experiments demonstrate the superior performance of our Recurrent Spiking Transformer framework in comparison to other spike neural network-based methods. Our framework exhibits a substantial margin of improvement in capturing and highlighting visual saliency in the spike stream, which not only provides a new perspective for spike-based saliency segmentation but also shows a new paradigm for full SNN-based transformer models. The code and dataset are available at https://github.com/BIT-Vision/SVS.
Recently, the task of Video Frame Prediction (VFP), which predicts future video frames from previous ones through extrapolation, has made remarkable progress. However, the performance of existing VFP methods is still far from satisfactory due to the fixed framerate video used: 1) they have difficulties in handling complex dynamic scenes; 2) they cannot predict future frames with flexible prediction time intervals. The event cameras can record the intensity changes asynchronously with a very high temporal resolution, which provides rich dynamic information about the observed scenes. In this paper, we propose to predict video frames from a single image and the following events, which can not only handle complex dynamic scenes but also predict future frames with flexible prediction time intervals. First, we introduce a symmetrical cross-modal attention augmentation module to enhance the complementary information between images and events. Second, we propose to jointly achieve optical flow estimation and frame generation by combining the motion information of events and the semantic information of the image, then inpainting the holes produced by forward warping to obtain an ideal prediction frame. Based on these, we propose a lightweight pyramidal coarse-to-fine model that can predict a 720P frame within 25 ms. Extensive experiments show that our proposed model significantly outperforms the state-of-the-art frame-based and event-based VFP methods and has the fastest runtime. Code is available at https://npucvr.github.io/VFPSIE/.
This study introduces a new method for detecting and localizing image forgery by focusing on manipulation traces within the noise domain. We posit that nearly invisible noise in RGB images carries tampering traces, useful for distinguishing and locating forgeries. However, the advancement of tampering technology complicates the direct application of noise for forgery detection, as the noise inconsistency between forged and authentic regions is not fully exploited. To tackle this, we develop a two-step discriminative noise-guided approach to explicitly enhance the representation and use of noise inconsistencies, thereby fully exploiting noise information to improve the accuracy and robustness of forgery detection. Specifically, we first enhance the noise discriminability of forged regions compared to authentic ones using a de-noising network and a statistics-based constraint. Then, we merge a model-driven guided filtering mechanism with a data-driven attention mechanism to create a learnable and differentiable noise-guided filter. This sophisticated filter allows us to maintain the edges of forged regions learned from the noise. Comprehensive experiments on multiple datasets demonstrate that our method can reliably detect and localize forgeries, surpassing existing state-of-the-art methods.
Free-hand sketches are appealing for humans as a universal tool to depict the visual world. Humans can recognize varied sketches of a category easily by identifying the concurrence and layout of the intrinsic semantic components of the category, since humans draw free-hand sketches based a common consensus that which types of semantic components constitute each sketch category. For example, an airplane should at least have a fuselage and wings. Based on this analysis, a semantic component-level memory module is constructed and embedded in the proposed structured sketch recognition network in this paper. The memory keys representing semantic components of each sketch category can be self-learned and enhance the recognition network's explainability. Our proposed networks can deal with different situations of sketch recognition, i.e., with or without semantic components labels of strokes. Experiments on the SPG and SketchIME datasets demonstrate the memory module's flexibility and the recognition network's explainability. The code and data are available at https://github.com/GuangmingZhu/SketchESC.
In this paper, we present SasWOT, the first training-free Semantic segmentation Architecture Search (SAS) framework via an auto-discovery proxy. Semantic segmentation is widely used in many real-time applications. For fast inference and memory efficiency, Previous SAS seeks the optimal segmenter by differentiable or RL Search. However, the significant computational costs of these training-based SAS limit their practical usage. To improve the search efficiency, we explore the training-free route but empirically observe that the existing zero-cost proxies designed on the classification task are sub-optimal on the segmentation benchmark. To address this challenge, we develop a customized proxy search framework for SAS tasks to augment its predictive capabilities. Specifically, we design the proxy search space based on the some observations: (1) different inputs of segmenter statistics can be well combined; (2) some basic operators can effectively improve the correlation. Thus, we build computational graphs with multiple statistics as inputs and different advanced basis arithmetic as the primary operations to represent candidate proxies. Then, we employ an evolutionary algorithm to crossover and mutate the superior candidates in the population based on correlation evaluation. Finally, based on the searched proxy, we perform the segmenter search without candidate training. In this way, SasWOT not only enables automated proxy optimization for SAS tasks but also achieves significant search acceleration before the retrain stage. Extensive experiments on Cityscapes and CamVid datasets demonstrate that SasWOT achieves superior trade-off between accuracy and speed over several state-of-the-art techniques. More remarkably, on Cityscapes dataset, SasWOT achieves the performance of 71.3% mIoU with the speed of 162 FPS.
Human intention understanding in untrimmed videos aims to watch a natural video and predict what the person’s intention is. Currently, exploration of predicting human intentions in untrimmed videos is far from enough. On the one hand, untrimmed videos with mixed actions and backgrounds have a significant long-tail distribution with concept drift characteristics. On the other hand, most methods can only perceive instantaneous intentions, but cannot determine the evolution of intentions. To solve the above challenges, we propose a loss based on Instance Confidence and Class Accuracy (ICCA), which aims to alleviate the prediction bias caused by the long-tail distribution with concept drift characteristics in video streams. In addition, we propose an intention-oriented evolutionary learning method to determine the intention evolution pattern (from what action to what action) and the time of evolution (when the action evolves). We conducted extensive experiments on two untrimmed video datasets (THUMOS14 and ActivityNET v1.3), and our method has achieved excellent results compared to SOTA methods. The code and supplementary materials are available at https://github.com/Jennifer123www/UntrimmedVideo.
Transformer has shown outstanding performance on image denoising, but the existing Transformer methods for image denoising are with large model sizes and high computational complexity, which is unfriendly to resource-constrained devices. In this paper, we propose a Lightweight Image Denoising Transformer method (LIDFormer) based on Triple Multi-Dconv Head Transposed Attention (TMDTA) to boost computational efficiency. LIDFormer first implements Discrete Wavelet Transform (DWT), which transforms the input image into a low-frequency space, greatly reducing the computational complexity of image denoising. However, the low-frequency image lacks fine-feature information, which degrades the denoising performance. To handle this problem, we introduce the Complementary Periodic Feature Reusing (CPFR) scheme for aggregating the shallow-layer features and the deep-layer features. Furthermore, TMDTA is proposed to integrate global context along three dimensions, thereby enhancing the ability of global feature representation. Note that our method can be applied as a pipeline for both convolutional neural networks and Transformers. Extensive experiments on several benchmarks demonstrate that the proposed LIDFormer achieves a better trade-off between high performance and low computational complexity on real-world image denoising tasks.
Optical Flow Estimation aims to find the 2D dense motion field between two frames. Due to the limitation of model structures and training datasets, existing methods often rely too much on local clues and ignore the integrity of objects, resulting in fragmented motion estimation. Through theoretical analysis, we find the pre-trained large vision models are helpful in optical flow estimation, and we notice that the recently famous Segment Anything Model (SAM) demonstrates a strong ability to segment complete objects, which is suitable for solving the fragmentation problem. We thus propose a solution to embed the frozen SAM image encoder into FlowFormer to enhance object perception. To address the challenge of in-depth utilizing SAM in non-segmentation tasks like optical flow estimation, we propose an Optical Flow Task-Specific Adaption scheme, including a Context Fusion Module to fuse the SAM encoder with the optical flow context encoder, and a Context Adaption Module to adapt the SAM features for optical flow task with Learned Task-Specific Embedding. Our proposed SAMFlow model reaches 0.86/2.10 clean/final EPE and 3.55/12.32 EPE/F1-all on Sintel and KITTI-15 training set, surpassing Flowformer by 8.5%/9.9% and 13.2%/16.3%. Furthermore, our model achieves state-of-the-art performance on the Sintel and KITTI-15 benchmarks, ranking #1 among all two-frame methods on Sintel clean pass.
Denoising Diffusion models have exhibited remarkable capabilities in image generation. However, generating high-quality samples requires a large number of iterations. Knowledge distillation for diffusion models is an effective method to address this limitation with a shortened sampling process but causes degraded generative quality. Based on our analysis with bias-variance decomposition and experimental observations, we attribute the degradation to the spatial fitting error occurring in the training of both the teacher and student model in the distillation. Accordingly, we propose Spatial Fitting-Error Reduction Distillation model (SFERD). SFERD utilizes attention guidance from the teacher model and a designed semantic gradient predictor to reduce the student's fitting error. Empirically, our proposed model facilitates high-quality sample generation in a few function evaluations. We achieve an FID of 5.31 on CIFAR-10 and 9.39 on ImageNet 64x64 with only one step, outperforming existing diffusion methods. Our study provides a new perspective on diffusion distillation by highlighting the intrinsic denoising ability of models.
Image registration plays a crucial role in histological image analysis, encompassing tasks like multi-modality fusion and disease grading. Traditional registration methods optimize objective functions for each image pair, yielding reliable accuracy but demanding heavy inference burdens. Recently, learning-based registration methods utilize networks to learn the optimization process during training and apply a one-step forward process during testing. While these methods offer promising registration performance with reduced inference time, they remain sensitive to appearance variances and local structure changes commonly encountered in histological image registration scenarios. In this paper, for the first time, we propose a novel test-time adaptation method for histological image registration, aiming to improve the generalization ability of learning-based methods. Specifically, we design two operations, style guidance and shape guidance, for the test-time adaptation process. The former leverages style representations encoded by feature statistics to address the issue of appearance variances, while the latter incorporates shape representations encoded by HOG features to improve registration accuracy in regions with structural changes. Furthermore, we consider the continuity of the model during the test-time adaptation process. Different from the previous methods initialized by a given trained model, we introduce a smoothing strategy to leverage historical models for better generalization. We conduct experiments with several representative learning-based backbones on the public histological dataset, demonstrating the superior registration performance of our test-time adaptation method.