Knowledge distillation (KD) aims to improve the performance of lightweight student networks under the guidance of pre-trained teachers. However, the large capacity gap between teachers and students limits the distillation gains. Previous methods addressing this problem have two weaknesses. First, most of them decrease the performance of pre-trained teachers, hindering students from achieving comparable performance. Second, these methods fail to dynamically adjust the transferred knowledge to be compatible with the representation ability of students, which is less effective in bridging the capacity gap. In this paper, we propose Adaptive Dual Guidance Knowledge Distillation (ADG-KD), which retains the guidance of the pre-trained teacher and uses the teacher's bidirectional optimization route guiding the student to alleviate the capacity gap problem. Specifically, ADG-KD introduces an initialized teacher, which has an identical structure to the pre-trained teacher and is optimized through the bidirectional supervision from both the pre-trained teacher and student. In this way, we construct the teacher's bidirectional optimization route to provide the students with an easy-to-hard and compatible knowledge sequence. ADG-KD trains the students under the proposed dual guidance approaches and automatically determines their importance weights, making the transferred knowledge better compatible with the representation ability of students. Extensive experiments on CIFAR-100, ImageNet, and MS-COCO demonstrate the effectiveness of our method.

This study focuses on multi-human relief modeling using a single photograph. Although previous studies successfully modeled 3D humans from single photographs, they were limited to reconstructing 3D individuals and could not be applied to multi-human scenes with complex inter-body and outer-body occlusions. In this study, we introduce MMRelief, a novel solution that takes a significant step toward high-quality and generalized multi-human relief modeling. MMRelief uses a three-step approach to achieve its objectives. First, it predicts an occlusion-aware depth map based on ZoeDepth [12]. Subsequently, it predicts a detailed normal map using a photo-to-normal network. Finally, MMRelief combines the strengths of both maps and constructs human relief using depth-constrained normal integration. Experimental results demonstrate that MMRelief has achieved state-of-the-art performance in normal human estimation. It can handle different styles of human photos with varying poses and dresses while producing reliefs with accurate body occlusions, reasonable depth ordering, and faithful geometrical details. The project page is at https://github.com/yanqingliu3856/MMRelief.

Abstract Gamma spectrum denoising can reduce the adverse effects of statistical fluctuations of radioactivity, gamma ray scattering, and electronic noise on the measured gamma spectrum. Traditional denoising methods are intricate and require analytical expertise in gamma spectrum analysis. This paper proposes a segmentation‐enhanced Convolutional Neural Network‐Stacked Denoising Autoencoder (CNN‐SDAE) method based on convolutional feature extraction network and stacked denoising autoencoder to achieve gamma spectrum denoising, which adopts the idea of data segmentation to enhance the learning ability of the neural network. By dividing the complete gamma spectrum into multiple segments and then using the segmentation‐enhanced CNN‐SDAE method for denoising, the method can achieve adaptive denoising without manually setting the threshold. The experimental results show that our method can effectively achieve gamma spectrum denoising while retaining the characteristics of the gamma spectrum. Compared with traditional methods, the denoising speed and effectiveness have been significantly improved, and the proposed method demonstrates an approximately 1.72‐fold enhancement in smoothing performance than the empirical mode decomposition method. Furthermore, in terms of retaining gamma spectrum characteristics, it also achieves a performance improvement of approximately three orders of magnitude than the wavelet method.

This paper aims at extending the method of Zhang et al. (2023) to produce not only portrait bas-reliefs from single photographs, but also high-depth reliefs with reasonable depth ordering. We cast this task as a problem of style-aware photo-to-depth translation, where the input is a photograph conditioned by a style vector and the output is a portrait relief with desired depth style. To construct ground-truth data for network training, we first propose an optimization-based method to synthesize high-depth reliefs from 3D portraits. Then, we train a normal-to-depth network to learn the mapping from normal maps to relief depths. After that, we use the trained network to generate high-depth relief samples using the provided normal maps from Zhang et al. (2023). As each normal map has pixel-wise photograph, we are able to establish correspondences between photographs and high-depth reliefs. By taking the bas-reliefs of Zhang et al. (2023), the new high-depth reliefs and their mixtures as target ground-truths, we finally train a encoder-to-decoder network to achieve style-aware relief modeling. Specially, the network is based on a U-shaped architecture, consisting of Swin Transformer blocks to process hierarchical deep features. Extensive experiments have demonstrated the effectiveness of the proposed method. Comparisons with previous works have verified its flexibility and state-of-the-art performance.

With the wide applications of biomedical images in the medical field, the segmentation of biomedical images plays an important role in clinical diagnosis, pathological analysis, and medical intervention. Full convolutional neural networks, especially U‐net, have improved the performance of segmentation greatly in recent years. However, due to their regular geometric structure, the standard convolutions that they use are inherently limited in dealing with geometric transformations while biomedical objects have huge variations in shape and size. In this study, the authors propose the DRU‐net, which is a novel U‐net with deformable encoder and reshaping upsampling convolution decoder, for biomedical image segmentation. First, deformable convolutional networks are applied and improved to enhance the learning ability of the encoder for geometric transformations. Second, a novel upsampling method named reshape upsampling convolution is proposed for better‐restoring resolution and fusion features. Furthermore, focal loss is used to address class imbalance and model overwhelmed problems in biomedical image segmentation tasks. Theoretic analysis and experimental results have shown that the proposed algorithm not only reduces the number of parameters of U‐Net, but also achieves produces competitive results compared with the state‐of‐the‐art algorithms in terms of various quantitative measures on Drosophila electron microscopy dataset and Warwick‐QU dataset.