In this paper we propose an improved mean-field inference algorithm for the fully connected paired CRFs model. The improved method Message Passing operation is changed from the original linear convolution to the present graph attention operation, while the process of the inference algorithm is turned into the forward process of the GAT model. Combined with the mean-field inferred label distribution, it is equivalent to the output of a classifier with only unary potential. To this end, we propose a graph attention network model with residual structure, and the model approach is applicable to all sequence annotation tasks, such as pixel-level image semantic segmentation tasks as well as text annotation tasks.
Shoe tread impressions are one of the most common types of evidence left at crime scenes. However, the utility of such evidence is limited by the lack of databases of footwear impression patterns that cover the huge and growing number of distinct shoe models. We propose to address this gap by leveraging shoe tread photographs collected by online retailers. The core challenge is to predict the impression pattern from the shoe photograph since ground-truth impressions or 3D shapes of tread patterns are not available. We develop a model that performs intrinsic image decomposition (predicting depth, normal, albedo, and lighting) from a single tread photo. Our approach, which we term ShoeRinsics, combines domain adaptation and re-rendering losses in order to leverage a mix of fully supervised synthetic data and unsupervised retail image data. To validate model performance, we also collected a set of paired shoe-sole images and corresponding prints, and define a benchmarking protocol to quantify the accuracy of predicted impressions. On this benchmark, ShoeRinsics outperforms existing methods for depth prediction and synthetic-to-real domain adaptation.
Deep learning based approaches has achieved great performance in single image super-resolution (SISR). However, recent advances in efficient super-resolution focus on reducing the number of parameters and FLOPs, and they aggregate more powerful features by improving feature utilization through complex layer connection strategies. These structures may not be necessary to achieve higher running speed, which makes them difficult to be deployed to resource-constrained devices. In this work, we propose a novel Residual Local Feature Network (RLFN). The main idea is using three convolutional layers for residual local feature learning to simplify feature aggregation, which achieves a good trade-off between model performance and inference time. Moreover, we revisit the popular contrastive loss and observe that the selection of intermediate features of its feature extractor has great influence on the performance. Besides, we propose a novel multi-stage warm-start training strategy. In each stage, the pre-trained weights from previous stages are utilized to improve the model performance. Combined with the improved contrastive loss and training strategy, the proposed RLFN outperforms all the state-of-the-art efficient image SR models in terms of runtime while maintaining both PSNR and SSIM for SR. In addition, we won the first place in the runtime track of the NTIRE 2022 efficient super-resolution challenge. Code will be available at https://github.com/fyan111/RLFN.
Image relighting aims to recalibrate the illumination setting in an image. In this paper, we propose a deep learning-based method called multi-modal bifurcated network (MBNet) for depth guided image relighting. That is, given an image and the corresponding depth maps, a new image with the given illuminant angle and color temperature is generated by our network. This model extracts the image and the depth features by the bifurcated network in the encoder. To use the two features effectively, we adopt the dynamic dilated pyramid modules in the decoder. Moreover, to increase the variety of training data, we propose a novel data process pipeline to increase the number of the training data. Experiments conducted on the VIDIT dataset show that the proposed solution obtains the \textbf{1}$^{st}$ place in terms of SSIM and PMS in the NTIRE 2021 Depth Guide One-to-one Relighting Challenge.
Face parsing assigns pixel-wise semantic labels as the face representation for computers, which is the fundamental part of many advanced face technologies. Compared with 2D face parsing, 3D face parsing shows more potential to achieve better performance and further application, but it is still challenging due to 3D mesh data computation. Recent works introduced different methods for 3D surface segmentation, while the performance is still limited. In this paper, we propose a method based on the "3D-2D-3D" strategy to accomplish 3D face parsing. The topological disk-like 2D face image containing spatial and textural information is transformed from the sampled 3D face data through the face parameterization algorithm, and a specific 2D network called CPFNet is proposed to achieve the semantic segmentation of the 2D parameterized face data with multi-scale technologies and feature aggregation. The 2D semantic result is then inversely re-mapped to 3D face data, which finally achieves the 3D face parsing. Experimental results show that both CPFNet and the "3D-2D-3D" strategy accomplish high-quality 3D face parsing and outperform state-of-the-art 2D networks as well as 3D methods in both qualitative and quantitative comparisons.
Unmanned Aerial Vehicles (UAVs) based video text spotting has been extensively used in civil and military domains. UAV's limited battery capacity motivates us to develop an energy-efficient video text spotting solution. In this paper, we first revisit RCNN's crop & resize training strategy and empirically find that it outperforms aligned RoI sampling on a real-world video text dataset captured by UAV. To reduce energy consumption, we further propose a multi-stage image processor that takes videos' redundancy, continuity, and mixed degradation into account. Lastly, the model is pruned and quantized before deployed on Raspberry Pi. Our proposed energy-efficient video text spotting solution, dubbed as E^2VTS, outperforms all previous methods by achieving a competitive tradeoff between energy efficiency and performance. All our codes and pre-trained models are available at https://github.com/wuzhenyusjtu/LPCVC20-VideoTextSpotting.
A 3D scene consists of a set of objects, each with a shape and a layout giving their position in space. Understanding 3D scenes from 2D images is an important goal, with applications in robotics and graphics. While there have been recent advances in predicting 3D shape and layout from a single image, most approaches rely on 3D ground truth for training which is expensive to collect at scale. We overcome these limitations and propose a method that learns to predict 3D shape and layout for objects without any ground truth shape or layout information: instead we rely on multi-view images with 2D supervision which can more easily be collected at scale. Through extensive experiments on 3D Warehouse, Hypersim, and ScanNet we demonstrate that our approach scales to large datasets of realistic images, and compares favorably to methods relying on 3D ground truth. On Hypersim and ScanNet where reliable 3D ground truth is not available, our approach outperforms supervised approaches trained on smaller and less diverse datasets.
Deep learning models have been deployed in numerous real-world applications such as autonomous driving and surveillance. However, these models are vulnerable in adversarial environments. Backdoor attack is emerging as a severe security threat which injects a backdoor trigger into a small portion of training data such that the trained model behaves normally on benign inputs but gives incorrect predictions when the specific trigger appears. While most research in backdoor attacks focuses on image classification, backdoor attacks on object detection have not been explored but are of equal importance. Object detection has been adopted as an important module in various security-sensitive applications such as autonomous driving. Therefore, backdoor attacks on object detection could pose severe threats to human lives and properties. We propose four kinds of backdoor attacks for object detection task: 1) Object Generation Attack: a trigger can falsely generate an object of the target class; 2) Regional Misclassification Attack: a trigger can change the prediction of a surrounding object to the target class; 3) Global Misclassification Attack: a single trigger can change the predictions of all objects in an image to the target class; and 4) Object Disappearance Attack: a trigger can make the detector fail to detect the object of the target class. We develop appropriate metrics to evaluate the four backdoor attacks on object detection. We perform experiments using two typical object detection models -- Faster-RCNN and YOLOv3 on different datasets. More crucially, we demonstrate that even fine-tuning on another benign dataset cannot remove the backdoor hidden in the object detection model. To defend against these backdoor attacks, we propose Detector Cleanse, an entropy-based run-time detection framework to identify poisoned testing samples for any deployed object detector.
In large-scale image retrieval, many indexing methods have been proposed to narrow down the searching scope of retrieval. The features extracted from images usually are of high dimensions or unfixed sizes due to the existence of key points. Most of existing index structures suffer from the dimension curse, the unfixed feature size and/or the loss of semantic similarity. In this paper a new classification-based indexing structure, called Semantic Indexing Structure (SIS), is proposed, in which we utilize the semantic categories rather than clustering centers to create database partitions, such that the proposed index SIS can be combined with feature extractors without the restriction of dimensions. Besides, it is observed that the size of each semantic partition is positively correlated with the semantic distribution of database. Along this way, we found that when the partition number is normalized to five, the proposed algorithm performed very well in all the tests. Compared with state-of-the-art models, SIS achieves outstanding performance.
Pancreatic cancer is one of the most malignant cancers in the world, which deteriorates rapidly with very high mortality. The rapid on-site evaluation (ROSE) technique innovates the workflow by immediately analyzing the fast stained cytopathological images with on-site pathologists, which enables faster diagnosis in this time-pressured process. However, the wider expansion of ROSE diagnosis has been hindered by the lack of experienced pathologists. To overcome this problem, we propose a hybrid high-performance deep learning model to enable the automated workflow, thus freeing the occupation of the valuable time of pathologists. By firstly introducing the Transformer block into this field with our particular multi-stage hybrid design, the spatial features generated by the convolutional neural network (CNN) significantly enhance the Transformer global modeling. Turning multi-stage spatial features as global attention guidance, this design combines the robustness from the inductive bias of CNN with the sophisticated global modeling power of Transformer. A dataset of 4240 ROSE images is collected to evaluate the method in this unexplored field. The proposed multi-stage hybrid Transformer (MSHT) achieves 95.68% in classification accuracy, which is distinctively higher than the state-of-the-art models. Facing the need for interpretability, MSHT outperforms its counterparts with more accurate attention regions. The results demonstrate that the MSHT can distinguish cancer samples accurately at an unprecedented image scale, laying the foundation for deploying automatic decision systems and enabling the expansion of ROSE in clinical practice. The code and records are available at: https://github.com/sagizty/Multi-Stage-Hybrid-Transformer.