论文检索

Boming Zhao, Luwei Yang, Mao Mao, Hujun Bao, Zhaopeng Cui

Due to the ability to synthesize high-quality novel views, Neural Radiance Fields (NeRF) has been recently exploited to improve visual localization in a known environment. However, the existing methods mostly utilize NeRF for data augmentation to improve the regression model training, and their performances on novel viewpoints and appearances are still limited due to the lack of geometric constraints. In this paper, we propose a novel visual localization framework, i.e., PNeRFLoc, based on a unified point-based representation. On one hand, PNeRFLoc supports the initial pose estimation by matching 2D and 3D feature points as traditional structure-based methods; on the other hand, it also enables pose refinement with novel view synthesis using rendering-based optimization. Specifically, we propose a novel feature adaption module to close the gaps between the features for visual localization and neural rendering. To improve the efficacy and efficiency of neural rendering-based optimization, we also developed an efficient rendering-based framework with a warping loss function. Extensive experiments demonstrate that PNeRFLoc performs the best on the synthetic dataset when the 3D NeRF model can be well learned, and significantly outperforms all the NeRF-boosted localization methods with on-par SOTA performance on the real-world benchmark localization datasets. Project webpage: https://zju3dv.github.io/PNeRFLoc/.

Zhenfei Zhang, Mingyang Li, Ming-Ching Chang

The ability to detect manipulation in multimedia data is vital in digital forensics. Existing Image Manipulation Detection (IMD) methods are mainly based on detecting anomalous features arisen from image editing or double compression artifacts. All existing IMD techniques encounter challenges when it comes to detecting small tampered regions from a large image. Moreover, compression-based IMD approaches face difficulties in cases of double compression of identical quality factors. To investigate the State-of-The-Art (SoTA) IMD methods in those challenging conditions, we introduce a new Challenging Image Manipulation Detection (CIMD) benchmark dataset, which consists of two subsets, for evaluating editing-based and compression-based IMD methods, respectively. The dataset images were manually taken and tampered with high-quality annotations. In addition, we propose a new two-branch network model based on HRNet that can better detect both the image-editing and compression artifacts in those challenging conditions. Extensive experiments on the CIMD benchmark show that our model significantly outperforms SoTA IMD methods on CIMD. The dataset is available at: https://github.com/ZhenfeiZ/CIMD.

Yin Zhang, Yongqiang Zhang, Zian Zhang, Man Zhang, Rui Tian, Mingli Ding

Object detection in dark conditions has always been a great challenge due to the complex formation process of low-light images. Currently, the mainstream methods usually adopt domain adaptation with Teacher-Student architecture to solve the dark object detection problem, and they imitate the dark conditions by using non-learnable data augmentation strategies on the annotated source daytime images. Note that these methods neglected to model the intrinsic imaging process, i.e. image signal processing (ISP), which is important for camera sensors to generate low-light images. To solve the above problems, in this paper, we propose a novel method named ISP-Teacher for dark object detection by exploring Teacher-Student architecture from a new perspective (i.e. self-supervised learning based ISP degradation). Specifically, we first design a day-to-night transformation module that consistent with the ISP pipeline of the camera sensors (ISP-DTM) to make the augmented images look more in line with the natural low-light images captured by cameras, and the ISP-related parameters are learned in a self-supervised manner. Moreover, to avoid the conflict between the ISP degradation and detection tasks in a shared encoder, we propose a disentanglement regularization (DR) that minimizes the absolute value of cosine similarity to disentangle two tasks and push two gradients vectors as orthogonal as possible. Extensive experiments conducted on two benchmarks show the effectiveness of our method in dark object detection. In particular, ISP-Teacher achieves an improvement of +2.4% AP and +3.3% AP over the SOTA method on BDD100k and SHIFT datasets, respectively. The code can be found at https://github.com/zhangyin1996/ISP-Teacher.

Jing Zhang, Xiaoqiang Liu, Mingzhe Chen, Zhe Wang

Few-shot Visual Question Answering (VQA) realizes few-shot cross-modal learning, which is an emerging and challenging task in computer vision. Currently, most of the few-shot VQA methods are confined to simply extending few-shot classification methods to cross-modal tasks while ignoring the spatial distribution properties of multimodal features and cross-modal information interaction. To address this problem, we propose a novel Cross-modal feature Distribution Calibration Inference Network (CDCIN) in this paper, where a new concept named visual information entropy is proposed to realize multimodal features distribution calibration by cross-modal information interaction for more effective few-shot VQA. Visual information entropy is a statistical variable that represents the spatial distribution of visual features guided by the question, which is aligned before and after the reasoning process to mitigate redundant information and improve multi-modal features by our proposed visual information entropy calibration module. To further enhance the inference ability of cross-modal features, we additionally propose a novel pre-training method, where the reasoning sub-network of CDCIN is pretrained on the base class in a VQA classification paradigm and fine-tuned on the few-shot VQA datasets. Extensive experiments demonstrate that our proposed CDCIN achieves excellent performance on few-shot VQA and outperforms state-of-the-art methods on three widely used benchmark datasets.

Haiming Zhang, Xu Yan, Dongfeng Bai, Jiantao Gao, Pan Wang, Bingbing Liu, Shuguang Cui, Zhen Li

3D occupancy prediction is an emerging task that aims to estimate the occupancy states and semantics of 3D scenes using multi-view images. However, image-based scene perception encounters significant challenges in achieving accurate prediction due to the absence of geometric priors. In this paper, we address this issue by exploring cross-modal knowledge distillation in this task, i.e., we leverage a stronger multi-modal model to guide the visual model during training. In practice, we observe that directly applying features or logits alignment, proposed and widely used in bird's-eye-view (BEV) perception, does not yield satisfactory results. To overcome this problem, we introduce RadOcc, a Rendering assisted distillation paradigm for 3D Occupancy prediction. By employing differentiable volume rendering, we generate depth and semantic maps in perspective views and propose two novel consistency criteria between the rendered outputs of teacher and student models. Specifically, the depth consistency loss aligns the termination distributions of the rendered rays, while the semantic consistency loss mimics the intra-segment similarity guided by vision foundation models (VLMs). Experimental results on the nuScenes dataset demonstrate the effectiveness of our proposed method in improving various 3D occupancy prediction approaches, e.g., our proposed methodology enhances our baseline by 2.2% in the metric of mIoU and achieves 50% in Occ3D benchmark.

Bowen Zhang, Qing Liu, Jianming Zhang, Yilin Wang, Liyang Liu, Zhe Lin, Yifan Liu

Amodal scene analysis entails interpreting the occlusion relationship among scene elements and inferring the possible shapes of the invisible parts. Existing methods typically frame this task as an extended instance segmentation or a pair-wise object de-occlusion problem. In this work, we propose a new framework, which comprises a Holistic Occlusion Relation Inference (HORI) module followed by an instance-level Generative Mask Completion (GMC) module. Unlike previous approaches, which rely on mask completion results for occlusion reasoning, our HORI module directly predicts an occlusion relation matrix in a single pass. This approach is much more efficient than the pair-wise de-occlusion process and it naturally handles mutual occlusion, a common but often neglected situation. Moreover, we formulate the mask completion task as a generative process and use a diffusion-based GMC module for instance-level mask completion. This improves mask completion quality and provides multiple plausible solutions. We further introduce a large-scale amodal segmentation dataset with high-quality human annotations, including mutual occlusions. Experiments on our dataset and two public benchmarks demonstrate the advantages of our method. code public available at https://github.com/zbwxp/Amodal-AAAI.

Yang Zhan, Yuan Yuan, Zhitong Xiong

We introduce a novel task of 3D visual grounding in monocular RGB images using language descriptions with both appearance and geometry information. Specifically, we build a large-scale dataset, Mono3DRefer, which contains 3D object targets with their corresponding geometric text descriptions, generated by ChatGPT and refined manually. To foster this task, we propose Mono3DVG-TR, an end-to-end transformer-based network, which takes advantage of both the appearance and geometry information in text embeddings for multi-modal learning and 3D object localization. Depth predictor is designed to explicitly learn geometry features. The dual text-guided adapter is proposed to refine multiscale visual and geometry features of the referred object. Based on depth-text-visual stacking attention, the decoder fuses object-level geometric cues and visual appearance into a learnable query. Comprehensive benchmarks and some insightful analyses are provided for Mono3DVG. Extensive comparisons and ablation studies show that our method significantly outperforms all baselines. The dataset and code will be released.

Zhenlong Yuan, Jiakai Cao, Zhaoxin Li, Hao Jiang, Zhaoqi Wang

In this paper, we introduce Segmentation-Driven Deformation Multi-View Stereo (SD-MVS), a method that can effectively tackle challenges in 3D reconstruction of textureless areas. We are the first to adopt the Segment Anything Model (SAM) to distinguish semantic instances in scenes and further leverage these constraints for pixelwise patch deformation on both matching cost and propagation. Concurrently, we propose a unique refinement strategy that combines spherical coordinates and gradient descent on normals and pixelwise search interval on depths, significantly improving the completeness of reconstructed 3D model. Furthermore, we adopt the Expectation-Maximization (EM) algorithm to alternately optimize the aggregate matching cost and hyperparameters, effectively mitigating the problem of parameters being excessively dependent on empirical tuning. Evaluations on the ETH3D high-resolution multi-view stereo benchmark and the Tanks and Temples dataset demonstrate that our method can achieve state-of-the-art results with less time consumption.

Hongtian Yu, Yunjie Tian, Qixiang Ye, Yunfan Liu

Vision Transformers (ViTs) have achieved remarkable success in computer vision tasks. However, their potential in rotation-sensitive scenarios has not been fully explored, and this limitation may be inherently attributed to the lack of spatial invariance in the data-forwarding process. In this study, we present a novel approach, termed Spatial Transform Decoupling (STD), providing a simple-yet-effective solution for oriented object detection with ViTs. Built upon stacked ViT blocks, STD utilizes separate network branches to predict the position, size, and angle of bounding boxes, effectively harnessing the spatial transform potential of ViTs in a divide-and-conquer fashion. Moreover, by aggregating cascaded activation masks (CAMs) computed upon the regressed parameters, STD gradually enhances features within regions of interest (RoIs), which complements the self-attention mechanism. Without bells and whistles, STD achieves state-of-the-art performance on the benchmark datasets including DOTA-v1.0 (82.24% mAP) and HRSC2016 (98.55% mAP), which demonstrates the effectiveness of the proposed method. Source code is available at https://github.com/yuhongtian17/Spatial-Transform-Decoupling.

Yunfan Ye, Kai Xu, Yuhang Huang, Renjiao Yi, Zhiping Cai

Limited by the encoder-decoder architecture, learning-based edge detectors usually have difficulty predicting edge maps that satisfy both correctness and crispness. With the recent success of the diffusion probabilistic model (DPM), we found it is especially suitable for accurate and crisp edge detection since the denoising process is directly applied to the original image size. Therefore, we propose the first diffusion model for the task of general edge detection, which we call DiffusionEdge. To avoid expensive computational resources while retaining the final performance, we apply DPM in the latent space and enable the classic cross-entropy loss which is uncertainty-aware in pixel level to directly optimize the parameters in latent space in a distillation manner. We also adopt a decoupled architecture to speed up the denoising process and propose a corresponding adaptive Fourier filter to adjust the latent features of specific frequencies. With all the technical designs, DiffusionEdge can be stably trained with limited resources, predicting crisp and accurate edge maps with much fewer augmentation strategies. Extensive experiments on four edge detection benchmarks demonstrate the superiority of DiffusionEdge both in correctness and crispness. On the NYUDv2 dataset, compared to the second best, we increase the ODS, OIS (without post-processing) and AC by 30.2%, 28.1% and 65.1%, respectively. Code: https://github.com/GuHuangAI/DiffusionEdge.

Jingwen Ye, Ruonan Yu, Songhua Liu, Xinchao Wang

Adversarial attacks constitute a notable threat to machine learning systems, given their potential to induce erroneous predictions and classifications. However, within real-world contexts, the essential specifics of the deployed model are frequently treated as a black box, consequently mitigating the vulnerability to such attacks. Thus, enhancing the transferability of the adversarial samples has become a crucial area of research, which heavily relies on selecting appropriate surrogate models. To address this challenge, we propose a novel approach that generates adversarial attacks in a mutual-modality optimization scheme. Our approach is accomplished by leveraging the pre-trained CLIP model. Firstly, we conduct a visual attack on the clean image that causes semantic perturbations on the aligned embedding space with the other textual modality. Then, we apply the corresponding defense on the textual modality by updating the prompts, which forces the re-matching on the perturbed embedding space. Finally, to enhance the attack transferability, we utilize the iterative training strategy on the visual attack and the textual defense, where the two processes optimize from each other. We evaluate our approach on several benchmark datasets and demonstrate that our mutual-modal attack strategy can effectively produce high-transferable attacks, which are stable regardless of the target networks. Our approach outperforms state-of-the-art attack methods and can be readily deployed as a plug-and-play solution.

Fulong Ye, Guang Liu, Xinya Wu, Ledell Wu

Large Text-to-Image(T2I) diffusion models have shown a remarkable capability to produce photorealistic and diverse images based on text inputs. However, existing works only support limited language input, e.g., English, Chinese, and Japanese, leaving users beyond these languages underserved and blocking the global expansion of T2I models. Therefore, this paper presents AltDiffusion, a novel multilingual T2I diffusion model that supports eighteen different languages. Specifically, we first train a multilingual text encoder based on the knowledge distillation. Then we plug it into a pretrained English-only diffusion model and train the model with a two-stage schema to enhance the multilingual capability, including concept alignment and quality improvement stage on a large-scale multilingual dataset. Furthermore, we introduce a new benchmark, which includes Multilingual-General-18(MG-18) and Multilingual-Cultural-18(MC-18) datasets, to evaluate the capabilities of T2I diffusion models for generating high-quality images and capturing culture-specific concepts in different languages. Experimental results on both MG-18 and MC-18 demonstrate that AltDiffusion outperforms current state-of-the-art T2I models, e.g., Stable Diffusion in multilingual understanding, especially with respect to culture-specific concepts, while still having comparable capability for generating high-quality images. All source code and checkpoints could be found in https://github.com/superhero-7/AltDiffuson.

Yiyang Yao, Peng Liu, Tiancheng Zhao, Qianqian Zhang, Jiajia Liao, Chunxin Fang, Kyusong Lee, Qing Wang

Object detection (OD) in computer vision has made significant progress in recent years, transitioning from closed-set labels to open-vocabulary detection (OVD) based on large-scale vision-language pre-training (VLP). However, current evaluation methods and datasets are limited to testing generalization over object types and referral expressions, which do not provide a systematic, fine-grained, and accurate benchmark of OVD models' abilities. In this paper, we propose a new benchmark named OVDEval, which includes 9 sub-tasks and introduces evaluations on commonsense knowledge, attribute understanding, position understanding, object relation comprehension, and more. The dataset is meticulously created to provide hard negatives that challenge models' true understanding of visual and linguistic input. Additionally, we identify a problem with the popular Average Precision (AP) metric when benchmarking models on these fine-grained label datasets and propose a new metric called Non-Maximum Suppression Average Precision (NMS-AP) to address this issue. Extensive experimental results show that existing top OVD models all fail on the new tasks except for simple object types, demonstrating the value of the proposed dataset in pinpointing the weakness of current OVD models and guiding future research. Furthermore, the proposed NMS-AP metric is verified by experiments to provide a much more truthful evaluation of OVD models, whereas traditional AP metrics yield deceptive results. Data is available at https://github.com/om-ai-lab/OVDEval

Bang Yang, Yong Dai, Xuxin Cheng, Yaowei Li, Asif Raza, Yuexian Zou

While vision-language pre-trained models (VL-PTMs) have advanced multimodal research in recent years, their mastery in a few languages like English restricts their applicability in broader communities. To this end, there is an increasing interest in developing multilingual VL models via a joint-learning setup, which, however, could be unrealistic due to expensive costs and data availability. In this work, we propose to extend VL-PTMs' language capacity by continual language learning (CLL), where a model needs to update its linguistic knowledge incrementally without suffering from catastrophic forgetting (CF). We begin our study by introducing a model dubbed CLL-CLIP, which builds upon CLIP, a prevailing VL-PTM that has acquired image-English text alignment. Specifically, CLL-CLIP contains an expandable token embedding layer to handle linguistic differences. It solely trains token embeddings to improve memory stability and is optimized under cross-modal and cross-lingual objectives to learn the alignment between images and multilingual texts. To alleviate CF raised by covariate shift and lexical overlap, we further propose a novel approach that ensures the identical distribution of all token embeddings during initialization and regularizes token embedding learning during training. We construct a CLL benchmark covering 36 languages based on MSCOCO and XM3600 datasets and then evaluate multilingual image-text retrieval performance. Extensive experiments verify the effectiveness of CLL-CLIP and show that our approach can boost CLL-CLIP, e.g., by 6.7% in text-to-image average Recall@1 on XM3600, and improve various state-of-the-art methods consistently. Our code and data are available at https://github.com/yangbang18/CLFM.

Guoli Yan, Zichun Zhong, Jing Hua

Despite achieving impressive improvement in accuracy, most existing monocular 3D human mesh reconstruction methods require large-scale 2D/3D ground-truths for supervision, which limits their applications on unlabeled in-the-wild data that is ubiquitous. To alleviate the reliance on 2D/3D ground-truths, we present a self-supervised 3D human pose and shape reconstruction framework that relies only on self-consistency between intermediate representations of images and projected 2D predictions. Specifically, we extract 2D joints and depth maps from monocular images as proxy inputs, which provides complementary clues to infer accurate 3D human meshes. Furthermore, to reduce the impacts from noisy and ambiguous inputs while better concentrate on the high-quality information, we design an uncertainty-aware module to automatically learn the reliability of the inputs at body-joint level based on the consistency between 2D joints and depth map. Experiments on benchmark datasets show that our approach outperforms other state-of-the-art methods at similar supervision levels.

Shiyu Xuan, Shiliang Zhang

Supervised Contrastive Loss (SCL) is popular in visual representation learning. Given an anchor image, SCL pulls two types of positive samples, i.e., its augmentation and other images from the same class together, while pushes negative images apart to optimize the learned embedding. In the scenario of long-tailed recognition, where the number of samples in each class is imbalanced, treating two types of positive samples equally leads to the biased optimization for intra-category distance. In addition, similarity relationship among negative samples, that are ignored by SCL, also presents meaningful semantic cues. To improve the performance on long-tailed recognition, this paper addresses those two issues of SCL by decoupling the training objective. Specifically, it decouples two types of positives in SCL and optimizes their relations toward different objectives to alleviate the influence of the imbalanced dataset. We further propose a patch-based self distillation to transfer knowledge from head to tail classes to relieve the under-representation of tail classes. It uses patch-based features to mine shared visual patterns among different instances and leverages a self distillation procedure to transfer such knowledge. Experiments on different long-tailed classification benchmarks demonstrate the superiority of our method. For instance, it achieves the 57.7% top-1 accuracy on the ImageNet-LT dataset. Combined with the ensemble-based method, the performance can be further boosted to 59.7%, which substantially outperforms many recent works. Our code will be released.

Sen Xu, Shikui Wei, Tao Ruan, Lixin Liao

Deep superpixel algorithms have made remarkable strides by substituting hand-crafted features with learnable ones. Nevertheless, we observe that existing deep superpixel methods, serving as mid-level representation operations, remain sensitive to the statistical properties (e.g., color distribution, high-level semantics) embedded within the training dataset. Consequently, learnable features exhibit constrained discriminative capability, resulting in unsatisfactory pixel grouping performance, particularly in untrainable application scenarios. To address this issue, we propose the Content Disentangle Superpixel (CDS) algorithm to selectively separate the invariant inter-pixel correlations and statistical properties, i.e., style noise. Specifically, We first construct auxiliary modalities that are homologous to the original RGB image but have substantial stylistic variations. Then, driven by mutual information, we propose the local-grid correlation alignment across modalities to reduce the distribution discrepancy of adaptively selected features and learn invariant inter-pixel correlations. Afterwards, we perform global-style mutual information minimization to enforce the separation of invariant content and train data styles. The experimental results on four benchmark datasets demonstrate the superiority of our approach to existing state-of-the-art methods, regarding boundary adherence, generalization, and efficiency. Code and pre-trained model are available at https://github.com/rookiie/CDSpixel.

Kezheng Xiong, Maoji Zheng, Qingshan Xu, Chenglu Wen, Siqi Shen, Cheng Wang

Point cloud registration, a fundamental task in 3D computer vision, has remained largely unexplored in cross-source point clouds and unstructured scenes. The primary challenges arise from noise, outliers, and variations in scale and density. However, neglected geometric natures of point clouds restricts the performance of current methods. In this paper, we propose a novel method termed SPEAL to leverage skeletal representations for effective learning of intrinsic topologies of point clouds, facilitating robust capture of geometric intricacy. Specifically, we design the Skeleton Extraction Module to extract skeleton points and skeletal features in an unsupervised manner, which is inherently robust to noise and density variances. Then, we propose the Skeleton-Aware GeoTransformer to encode high-level skeleton-aware features. It explicitly captures the topological natures and inter-point-cloud skeletal correlations with the noise-robust and density-invariant skeletal representations. Next, we introduce the Correspondence Dual-Sampler to facilitate correspondences by augmenting the correspondence set with skeletal correspondences. Furthermore, we construct a challenging novel cross-source point cloud dataset named KITTI CrossSource for benchmarking cross-source point cloud registration methods. Extensive quantitative and qualitative experiments are conducted to demonstrate our approach’s superiority and robustness on both cross-source and same-source datasets. To the best of our knowledge, our approach is the first to facilitate point cloud registration with skeletal geometric priors.

Zheng Xing, Weibing Zhao

Action segmentation serves as a pivotal component in comprehending videos, encompassing the learning of a sequence of semantically consistent action units known as actoms. Conventional methodologies tend to require a significant consumption of time for both training and learning phases. This paper introduces an innovative unsupervised framework for action segmentation in video, characterized by its fast learning capability and absence of mandatory training. The core idea involves splitting the video into distinct actoms, which are then merging together based on shared actions. The key challenge here is to prevent the inadvertent creation of singular actoms that attempt to represent multiple actions during the splitting phase. Additionally, it is crucial to avoid situations where actoms associated with the same action are incorrectly grouped into multiple clusters during the merging phase. In this paper, we present a method for calculating the similarity between adjacent frames under a subspace assumption. Then, we employ a local minimum searching procedure, which effectively splits the video into coherent actoms aligned with their semantic meaning and provides us an action segmentation proposal. Subsequently, we calculate a spatio-temporal similarity between actoms, followed by developing a merging process to merge actoms representing identical actions within the action segmentation proposals. Our approach is evaluated on four benchmark datasets, and the results demonstrate that our method achieves state-of-the-art performance. Besides, our method also achieves the optimal balance between accuracy and learning time when compared to existing unsupervised techniques. Code is available at https://github.com/y66y/SaM.

Xinyi Wu, Wentao Ma, Dan Guo, Tongqing Zhou, Shan Zhao, Zhiping Cai

Text-based Person Re-identification (T-ReID), which aims at retrieving a specific pedestrian image from a collection of images via text-based information, has received significant attention. However, previous research has overlooked a challenging yet practical form of T-ReID: dealing with image galleries mixed with occluded and inconsistent personal visuals, instead of ideal visuals with a full-body and clear view. Its major challenges lay in the insufficiency of benchmark datasets and the enlarged semantic gap incurred by arbitrary occlusions and modality gap between text description and visual representation of the target person. To alleviate these issues, we first design an Occlusion Generator (OGor) for the automatic generation of artificial occluded images from generic surveillance images. Then, a fine-granularity token selection mechanism is proposed to minimize the negative impact of occlusion for robust feature learning, and a novel multi-granularity contrastive consistency alignment framework is designed to leverage intra-/inter-granularity of visual-text representations for semantic alignment of occluded visuals and query texts. Experimental results demonstrate that our method exhibits superior performance. We believe this work could inspire the community to investigate more dedicated designs for implementing T-ReID in real-world scenarios. The source code is available at https://github.com/littlexinyi/MGCC.