Yun Ye

Zheng Zhu, Guan Huang, Jiankang Deng et al. · tsinghua

Face benchmarks empower the research community to train and evaluate high-performance face recognition systems. In this paper, we contribute a new million-scale recognition benchmark, containing uncurated 4M identities/260M faces (WebFace260M) and cleaned 2M identities/42M faces (WebFace42M) training data, as well as an elaborately designed time-constrained evaluation protocol. Firstly, we collect 4M name lists and download 260M faces from the Internet. Then, a Cleaning Automatically utilizing Self-Training (CAST) pipeline is devised to purify the tremendous WebFace260M, which is efficient and scalable. To the best of our knowledge, the cleaned WebFace42M is the largest public face recognition training set and we expect to close the data gap between academia and industry. Referring to practical deployments, Face Recognition Under Inference Time conStraint (FRUITS) protocol and a new test set with rich attributes are constructed. Besides, we gather a large-scale masked face sub-set for biometrics assessment under COVID-19. For a comprehensive evaluation of face matchers, three recognition tasks are performed under standard, masked and unbiased settings, respectively. Equipped with this benchmark, we delve into million-scale face recognition problems. A distributed framework is developed to train face recognition models efficiently without tampering with the performance. Enabled by WebFace42M, we reduce 40% failure rate on the challenging IJB-C set and rank 3rd among 430 entries on NIST-FRVT. Even 10% data (WebFace4M) shows superior performance compared with the public training sets. Furthermore, comprehensive baselines are established under the FRUITS-100/500/1000 milliseconds protocols. The proposed benchmark shows enormous potential on standard, masked and unbiased face recognition scenarios. Our WebFace260M website is https://www.face-benchmark.org.

Xiaofeng Wang, Zheng Zhu, Wenbo Xu et al.

Semantic occupancy perception is essential for autonomous driving, as automated vehicles require a fine-grained perception of the 3D urban structures. However, existing relevant benchmarks lack diversity in urban scenes, and they only evaluate front-view predictions. Towards a comprehensive benchmarking of surrounding perception algorithms, we propose OpenOccupancy, which is the first surrounding semantic occupancy perception benchmark. In the OpenOccupancy benchmark, we extend the large-scale nuScenes dataset with dense semantic occupancy annotations. Previous annotations rely on LiDAR points superimposition, where some occupancy labels are missed due to sparse LiDAR channels. To mitigate the problem, we introduce the Augmenting And Purifying (AAP) pipeline to ~2x densify the annotations, where ~4000 human hours are involved in the labeling process. Besides, camera-based, LiDAR-based and multi-modal baselines are established for the OpenOccupancy benchmark. Furthermore, considering the complexity of surrounding occupancy perception lies in the computational burden of high-resolution 3D predictions, we propose the Cascade Occupancy Network (CONet) to refine the coarse prediction, which relatively enhances the performance by ~30% than the baseline. We hope the OpenOccupancy benchmark will boost the development of surrounding occupancy perception algorithms.

Xiaofeng Wang, Zheng Zhu, Guan Huang et al.

Self-supervised monocular methods can efficiently learn depth information of weakly textured surfaces or reflective objects. However, the depth accuracy is limited due to the inherent ambiguity in monocular geometric modeling. In contrast, multi-frame depth estimation methods improve the depth accuracy thanks to the success of Multi-View Stereo (MVS), which directly makes use of geometric constraints. Unfortunately, MVS often suffers from texture-less regions, non-Lambertian surfaces, and moving objects, especially in real-world video sequences without known camera motion and depth supervision. Therefore, we propose MOVEDepth, which exploits the MOnocular cues and VElocity guidance to improve multi-frame Depth learning. Unlike existing methods that enforce consistency between MVS depth and monocular depth, MOVEDepth boosts multi-frame depth learning by directly addressing the inherent problems of MVS. The key of our approach is to utilize monocular depth as a geometric priority to construct MVS cost volume, and adjust depth candidates of cost volume under the guidance of predicted camera velocity. We further fuse monocular depth and MVS depth by learning uncertainty in the cost volume, which results in a robust depth estimation against ambiguity in multi-view geometry. Extensive experiments show MOVEDepth achieves state-of-the-art performance: Compared with Monodepth2 and PackNet, our method relatively improves the depth accuracy by 20\% and 19.8\% on the KITTI benchmark. MOVEDepth also generalizes to the more challenging DDAD benchmark, relatively outperforming ManyDepth by 7.2\%. The code is available at https://github.com/JeffWang987/MOVEDepth.

Xiaofeng Wang, Zheng Zhu, Yunpeng Zhang et al.

In recent years, vision-centric perception has flourished in various autonomous driving tasks, including 3D detection, semantic map construction, motion forecasting, and depth estimation. Nevertheless, the latency of vision-centric approaches is too high for practical deployment (e.g., most camera-based 3D detectors have a runtime greater than 300ms). To bridge the gap between ideal research and real-world applications, it is necessary to quantify the trade-off between performance and efficiency. Traditionally, autonomous-driving perception benchmarks perform the offline evaluation, neglecting the inference time delay. To mitigate the problem, we propose the Autonomous-driving StreAming Perception (ASAP) benchmark, which is the first benchmark to evaluate the online performance of vision-centric perception in autonomous driving. On the basis of the 2Hz annotated nuScenes dataset, we first propose an annotation-extending pipeline to generate high-frame-rate labels for the 12Hz raw images. Referring to the practical deployment, the Streaming Perception Under constRained-computation (SPUR) evaluation protocol is further constructed, where the 12Hz inputs are utilized for streaming evaluation under the constraints of different computational resources. In the ASAP benchmark, comprehensive experiment results reveal that the model rank alters under different constraints, suggesting that the model latency and computation budget should be considered as design choices to optimize the practical deployment. To facilitate further research, we establish baselines for camera-based streaming 3D detection, which consistently enhance the streaming performance across various hardware. ASAP project page: https://github.com/JeffWang987/ASAP.

Junjie Huang, Yun Ye, Zhujin Liang et al.

3D object Detection with LiDAR-camera encounters overfitting in algorithm development which is derived from the violation of some fundamental rules. We refer to the data annotation in dataset construction for theory complementing and argue that the regression task prediction should not involve the feature from the camera branch. By following the cutting-edge perspective of 'Detecting As Labeling', we propose a novel paradigm dubbed DAL. With the most classical elementary algorithms, a simple predicting pipeline is constructed by imitating the data annotation process. Then we train it in the simplest way to minimize its dependency and strengthen its portability. Though simple in construction and training, the proposed DAL paradigm not only substantially pushes the performance boundary but also provides a superior trade-off between speed and accuracy among all existing methods. With comprehensive superiority, DAL is an ideal baseline for both future work development and practical deployment. The code has been released to facilitate future work on https://github.com/HuangJunJie2017/BEVDet.

Xiaofeng Wang, Zheng Zhu, Fangbo Qin et al.

Learning-based Multi-View Stereo (MVS) methods warp source images into the reference camera frustum to form 3D volumes, which are fused as a cost volume to be regularized by subsequent networks. The fusing step plays a vital role in bridging 2D semantics and 3D spatial associations. However, previous methods utilize extra networks to learn 2D information as fusing cues, underusing 3D spatial correlations and bringing additional computation costs. Therefore, we present MVSTER, which leverages the proposed epipolar Transformer to learn both 2D semantics and 3D spatial associations efficiently. Specifically, the epipolar Transformer utilizes a detachable monocular depth estimator to enhance 2D semantics and uses cross-attention to construct data-dependent 3D associations along epipolar line. Additionally, MVSTER is built in a cascade structure, where entropy-regularized optimal transport is leveraged to propagate finer depth estimations in each stage. Extensive experiments show MVSTER achieves state-of-the-art reconstruction performance with significantly higher efficiency: Compared with MVSNet and CasMVSNet, our MVSTER achieves 34% and 14% relative improvements on the DTU benchmark, with 80% and 51% relative reductions in running time. MVSTER also ranks first on Tanks&Temples-Advanced among all published works. Code is released at https://github.com/JeffWang987.

Jiayu Zou, Zheng Zhu, Yun Ye et al.

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. The code will be available soon.

Yunpeng Zhang, Deheng Qian, Ding Li et al.

Modeling complicated interactions among the ego-vehicle, road agents, and map elements has been a crucial part for safety-critical autonomous driving. Previous works on end-to-end autonomous driving rely on the attention mechanism for handling heterogeneous interactions, which fails to capture the geometric priors and is also computationally intensive. In this paper, we propose the Interaction Scene Graph (ISG) as a unified method to model the interactions among the ego-vehicle, road agents, and map elements. With the representation of the ISG, the driving agents aggregate essential information from the most influential elements, including the road agents with potential collisions and the map elements to follow. Since a mass of unnecessary interactions are omitted, the more efficient scene-graph-based framework is able to focus on indispensable connections and leads to better performance. We evaluate the proposed method for end-to-end autonomous driving on the nuScenes dataset. Compared with strong baselines, our method significantly outperforms in the full-stack driving tasks, including perception, prediction, and planning. Code will be released at https://github.com/zhangyp15/GraphAD.

Junjie Huang, Guan Huang, Zheng Zhu et al.

Autonomous driving perceives its surroundings for decision making, which is one of the most complex scenarios in visual perception. The success of paradigm innovation in solving the 2D object detection task inspires us to seek an elegant, feasible, and scalable paradigm for fundamentally pushing the performance boundary in this area. To this end, we contribute the BEVDet paradigm in this paper. BEVDet performs 3D object detection in Bird-Eye-View (BEV), where most target values are defined and route planning can be handily performed. We merely reuse existing modules to build its framework but substantially develop its performance by constructing an exclusive data augmentation strategy and upgrading the Non-Maximum Suppression strategy. In the experiment, BEVDet offers an excellent trade-off between accuracy and time-efficiency. As a fast version, BEVDet-Tiny scores 31.2% mAP and 39.2% NDS on the nuScenes val set. It is comparable with FCOS3D, but requires just 11% computational budget of 215.3 GFLOPs and runs 9.2 times faster at 15.6 FPS. Another high-precision version dubbed BEVDet-Base scores 39.3% mAP and 47.2% NDS, significantly exceeding all published results. With a comparable inference speed, it surpasses FCOS3D by a large margin of +9.8% mAP and +10.0% NDS. The source code is publicly available for further research at https://github.com/HuangJunJie2017/BEVDet .

Yanyu Chen, Yao Yao, Wai Kin Victor Chan et al.

Data sparsity and cold-start problems are persistent challenges in recommendation systems. Cross-domain recommendation (CDR) is a promising solution that utilizes knowledge from the source domain to improve the recommendation performance in the target domain. Previous CDR approaches have mainly followed the Embedding and Mapping (EMCDR) framework, which involves learning a mapping function to facilitate knowledge transfer. However, these approaches necessitate re-engineering and re-training the network structure to incorporate transferrable knowledge, which can be computationally expensive and may result in catastrophic forgetting of the original knowledge. In this paper, we present a scalable and efficient paradigm to address data sparsity and cold-start issues in CDR, named CDR-Adapter, by decoupling the original recommendation model from the mapping function, without requiring re-engineering the network structure. Specifically, CDR-Adapter is a novel plug-and-play module that employs adapter modules to align feature representations, allowing for flexible knowledge transfer across different domains and efficient fine-tuning with minimal training costs. We conducted extensive experiments on the benchmark dataset, which demonstrated the effectiveness of our approach over several state-of-the-art CDR approaches.

Yun Ye, Yanjie Pan, Qually Jiang et al.

Over-fitting-based image compression requires weights compactness for compression and fast convergence for practical use, posing challenges for deep convolutional neural networks (CNNs) based methods. This paper presents a simple re-parameterization method to train CNNs with reduced weights storage and accelerated convergence. The convolution kernels are re-parameterized as a weighted sum of discrete cosine transform (DCT) kernels enabling direct optimization in the frequency domain. Combined with L1 regularization, the proposed method surpasses vanilla convolutions by achieving a significantly improved rate-distortion with low computational cost. The proposed method is verified with extensive experiments of over-fitting-based image restoration on various datasets, achieving up to -46.12% BD-rate on top of HEIF with only 200 iterations.

Zhehao Dong, Xiaofeng Wang, Zheng Zhu et al.

Vision-language-action (VLA) models increasingly rely on diverse training data to achieve robust generalization. However, collecting large-scale real-world robot manipulation data across varied object appearances and environmental conditions remains prohibitively time-consuming and expensive. To overcome this bottleneck, we propose Embodied Manipulation Media Adaptation (EMMA), a VLA policy enhancement framework that integrates a generative data engine with an effective training pipeline. We introduce DreamTransfer, a diffusion Transformer-based framework for generating multi-view consistent, geometrically grounded embodied manipulation videos. DreamTransfer enables text-controlled visual editing of robot videos, transforming foreground, background, and lighting conditions without compromising 3D structure or geometrical plausibility. Furthermore, we explore hybrid training with real and generated data, and introduce AdaMix, a hard-sample-aware training strategy that dynamically reweights training batches to focus optimization on perceptually or kinematically challenging samples. Extensive experiments show that videos generated by DreamTransfer significantly outperform prior video generation methods in multi-view consistency, geometric fidelity, and text-conditioning accuracy. Crucially, VLAs trained with generated data enable robots to generalize to unseen object categories and novel visual domains using only demonstrations from a single appearance. In real-world robotic manipulation tasks with zero-shot visual domains, our approach achieves over a 200% relative performance gain compared to training on real data alone, and further improves by 13% with AdaMix, demonstrating its effectiveness in boosting policy generalization.

GigaBrain Team, Angen Ye, Boyuan Wang et al.

Training Vision-Language-Action (VLA) models for generalist robots typically requires large-scale real-world robot data, which is expensive and time-consuming to collect. The inefficiency of physical data collection severely limits the scalability, and generalization capacity of current VLA systems. To address this challenge, we introduce GigaBrain-0, a novel VLA foundation model empowered by world model-generated data (e.g., video generation, real2real transfer, human transfer, view transfer, sim2real transfer data). By leveraging world models to generate diverse data at scale, GigaBrain-0 significantly reduces reliance on real robot data while improving cross-task generalization. Our approach further improves policy robustness through RGBD input modeling and embodied Chain-of-Thought (CoT) supervision, enabling the model to reason about spatial geometry, object states, and long-horizon dependencies during task execution. This leads to substantial gains in real-world performance on dexterous, long-horizon, and mobile manipulation tasks. Extensive experiments demonstrate that GigaBrain-0 achieves superior generalization across variations in appearances (e.g., textures, colors), object placements, and camera viewpoints. Additionally, we present GigaBrain-0-Small, an optimized lightweight variant designed to run efficiently on devices such as the NVIDIA Jetson AGX Orin.

Zheng Zhu, Guan Huang, Jiankang Deng et al.

According to WHO statistics, there are more than 204,617,027 confirmed COVID-19 cases including 4,323,247 deaths worldwide till August 12, 2021. During the coronavirus epidemic, almost everyone wears a facial mask. Traditionally, face recognition approaches process mostly non-occluded faces, which include primary facial features such as the eyes, nose, and mouth. Removing the mask for authentication in airports or laboratories will increase the risk of virus infection, posing a huge challenge to current face recognition systems. Due to the sudden outbreak of the epidemic, there are yet no publicly available real-world masked face recognition (MFR) benchmark. To cope with the above-mentioned issue, we organize the Face Bio-metrics under COVID Workshop and Masked Face Recognition Challenge in ICCV 2021. Enabled by the ultra-large-scale WebFace260M benchmark and the Face Recognition Under Inference Time conStraint (FRUITS) protocol, this challenge (WebFace260M Track) aims to push the frontiers of practical MFR. Since public evaluation sets are mostly saturated or contain noise, a new test set is gathered consisting of elaborated 2,478 celebrities and 60,926 faces. Meanwhile, we collect the world-largest real-world masked test set. In the first phase of WebFace260M Track, 69 teams (total 833 solutions) participate in the challenge and 49 teams exceed the performance of our baseline. There are second phase of the challenge till October 1, 2021 and on-going leaderboard. We will actively update this report in the future.

Zheng Zhu, Guan Huang, Jiankang Deng et al.

In this paper, we contribute a new million-scale face benchmark containing noisy 4M identities/260M faces (WebFace260M) and cleaned 2M identities/42M faces (WebFace42M) training data, as well as an elaborately designed time-constrained evaluation protocol. Firstly, we collect 4M name list and download 260M faces from the Internet. Then, a Cleaning Automatically utilizing Self-Training (CAST) pipeline is devised to purify the tremendous WebFace260M, which is efficient and scalable. To the best of our knowledge, the cleaned WebFace42M is the largest public face recognition training set and we expect to close the data gap between academia and industry. Referring to practical scenarios, Face Recognition Under Inference Time conStraint (FRUITS) protocol and a test set are constructed to comprehensively evaluate face matchers. Equipped with this benchmark, we delve into million-scale face recognition problems. A distributed framework is developed to train face recognition models efficiently without tampering with the performance. Empowered by WebFace42M, we reduce relative 40% failure rate on the challenging IJB-C set, and ranks the 3rd among 430 entries on NIST-FRVT. Even 10% data (WebFace4M) shows superior performance compared with public training set. Furthermore, comprehensive baselines are established on our rich-attribute test set under FRUITS-100ms/500ms/1000ms protocol, including MobileNet, EfficientNet, AttentionNet, ResNet, SENet, ResNeXt and RegNet families. Benchmark website is https://www.face-benchmark.org.

Qing Yang, Xia Zhu, Jong-Kae Fwu et al.

Deep neural networks (DNNs) have recently been applied and used in many advanced and diverse tasks, such as medical diagnosis, automatic driving, etc. Due to the lack of transparency of the deep models, DNNs are often criticized for their prediction that cannot be explainable by human. In this paper, we propose a novel Morphological Fragmental Perturbation Pyramid (MFPP) method to solve the Explainable AI problem. In particular, we focus on the black-box scheme, which can identify the input area that is responsible for the output of the DNN without having to understand the internal architecture of the DNN. In the MFPP method, we divide the input image into multi-scale fragments and randomly mask out fragments as perturbation to generate a saliency map, which indicates the significance of each pixel for the prediction result of the black box model. Compared with the existing input sampling perturbation method, the pyramid structure fragment has proved to be more effective. It can better explore the morphological information of the input image to match its semantic information, and does not need any value inside the DNN. We qualitatively and quantitatively prove that MFPP meets and exceeds the performance of state-of-the-art (SOTA) black-box interpretation method on multiple DNN models and datasets.

Qing Yang, Xia Zhu, Jong-Kae Fwu et al.

Face anti-spoofing has become an increasingly important and critical security feature for authentication systems, due to rampant and easily launchable presentation attacks. Addressing the shortage of multi-modal face dataset, CASIA recently released the largest up-to-date CASIA-SURF Cross-ethnicity Face Anti-spoofing(CeFA) dataset, covering 3 ethnicities, 3 modalities, 1607 subjects, and 2D plus 3D attack types in four protocols, and focusing on the challenge of improving the generalization capability of face anti-spoofing in cross-ethnicity and multi-modal continuous data. In this paper, we propose a novel pipeline-based multi-stream CNN architecture called PipeNet for multi-modal face anti-spoofing. Unlike previous works, Selective Modal Pipeline (SMP) is designed to enable a customized pipeline for each data modality to take full advantage of multi-modal data. Limited Frame Vote (LFV) is designed to ensure stable and accurate prediction for video classification. The proposed method wins the third place in the final ranking of Chalearn Multi-modal Cross-ethnicity Face Anti-spoofing Recognition Challenge@CVPR2020. Our final submission achieves the Average Classification Error Rate (ACER) of 2.21 with Standard Deviation of 1.26 on the test set.

Yun Ye, Ganmei You, Jong-Kae Fwu et al.

Structured pruning, especially channel pruning is widely used for the reduced computational cost and the compatibility with off-the-shelf hardware devices. Among existing works, weights are typically removed using a predefined global threshold, or a threshold computed from a predefined metric. The predefined global threshold based designs ignore the variation among different layers and weights distribution, therefore, they may often result in sub-optimal performance caused by over-pruning or under-pruning. In this paper, we present a simple yet effective method, termed Optimal Thresholding (OT), to prune channels with layer dependent thresholds that optimally separate important from negligible channels. By using OT, most negligible or unimportant channels are pruned to achieve high sparsity while minimizing performance degradation. Since most important weights are preserved, the pruned model can be further fine-tuned and quickly converge with very few iterations. Our method demonstrates superior performance, especially when compared to the state-of-the-art designs at high levels of sparsity. On CIFAR-100, a pruned and fine-tuned DenseNet-121 by using OT achieves 75.99% accuracy with only 1.46e8 FLOPs and 0.71M parameters.

Sun Yu, Ye Yun, Liu Wu et al.

We describe an end-to-end method for recovering 3D human body mesh from single images and monocular videos. Different from the existing methods try to obtain all the complex 3D pose, shape, and camera parameters from one coupling feature, we propose a skeleton-disentangling based framework, which divides this task into multi-level spatial and temporal granularity in a decoupling manner. In spatial, we propose an effective and pluggable "disentangling the skeleton from the details" (DSD) module. It reduces the complexity and decouples the skeleton, which lays a good foundation for temporal modeling. In temporal, the self-attention based temporal convolution network is proposed to efficiently exploit the short and long-term temporal cues. Furthermore, an unsupervised adversarial training strategy, temporal shuffles and order recovery, is designed to promote the learning of motion dynamics. The proposed method outperforms the state-of-the-art 3D human mesh recovery methods by 15.4% MPJPE and 23.8% PA-MPJPE on Human3.6M. State-of-the-art results are also achieved on the 3D pose in the wild (3DPW) dataset without any fine-tuning. Especially, ablation studies demonstrate that skeleton-disentangled representation is crucial for better temporal modeling and generalization.

Yun Ye, Yixin Li, Bo Wu et al.

Fashion attribute classification is of great importance to many high-level tasks such as fashion item search, fashion trend analysis, fashion recommendation, etc. The task is challenging due to the extremely imbalanced data distribution, particularly the attributes with only a few positive samples. In this paper, we introduce a hard-aware pipeline to make full use of "hard" samples/attributes. We first propose Hard-Aware BackPropagation (HABP) to efficiently and adaptively focus on training "hard" data. Then for the identified hard labels, we propose to synthesize more complementary samples for training. To stabilize training, we extend semi-supervised GAN by directly deactivating outputs for synthetic complementary samples (Deact). In general, our method is more effective in addressing "hard" cases. HABP weights more on "hard" samples. For "hard" attributes with insufficient training data, Deact brings more stable synthetic samples for training and further improve the performance. Our method is verified on large scale fashion dataset, outperforming other state-of-the-art without any additional supervisions.

Yixin Li, Shengqin Tang, Yun Ye et al.

Fashion landmark detection is a challenging task even using the current deep learning techniques, due to the large variation and non-rigid deformation of clothes. In order to tackle these problems, we propose Spatial-Aware Non-Local (SANL) block, an attentive module in deep neural network which can utilize spatial information while capturing global dependency. Actually, the SANL block is constructed from the non-local block in the residual manner which can learn the spatial related representation by taking a spatial attention map from Grad-CAM. We then establish our fashion landmark detection framework on feature pyramid network, equipped with four SANL blocks in the backbone. It is demonstrated by the experimental results on two large-scale fashion datasets that our proposed fashion landmark detection approach with the SANL blocks outperforms the current state-of-the-art methods considerably. Some supplementary experiments on fine-grained image classification also show the effectiveness of the proposed SANL block.

Junhui Wu, Yun Ye, Yu Chen et al.

In this paper, we propose a simple yet effective solution to a change detection task that detects the difference between two images, which we call "spot the difference". Our approach uses CNN-based object detection by stacking two aligned images as input and considering the differences between the two images as objects to detect. An early-merging architecture is used as the backbone network. Our method is accurate, fast and robust while using very cheap annotation. We verify the proposed method on the task of change detection between the digital design and its photographic image of a book. Compared to verification based methods, our object detection based method outperforms other methods by a large margin and gives extra information of location. We compress the network and achieve 24 times acceleration while keeping the accuracy. Besides, as we synthesize the training data for detection using weakly labeled images, our method does not need expensive bounding box annotation.