Current methods aggregate multilevel features from the backbone and introduce edge information to get more refined saliency maps. However, little attention is paid to how to suppress the regions with similar saliency appearances in the background. These regions usually exist in the vicinity of salient objects and have high contrast with the background, which is easy to be misclassified as foreground. To solve this problem, we propose a gated feature interaction network (GFINet) to integrate multiple saliency features, which can utilize nonboundary features with background information to suppress pseudosalient objects and simultaneously apply boundary features to supplement edge details. Different from previous methods that only consider the complementarity between saliency and boundary, the proposed network introduces nonboundary features into the decoder to filter the pseudosalient objects. Specifically, GFINet consists of global features aggregation branch (GFAB), boundary and nonboundary features' perception branch (B&NFPB), and gated feature interaction module (GFIM). According to the global features generated by GFAB, boundary and nonboundary features produced by B&NFPB and GFIM employ a gate structure to adaptively optimize the saliency information interchange between abovementioned features and, thus, predict the final saliency maps. Besides, due to the imbalanced distribution between the boundary pixels and nonboundary ones, the binary cross-entropy (BCE) loss is difficult to predict the pixels near the boundary. Therefore, we design a border region aware (BRA) loss to further boost the quality of boundary and nonboundary, which can guide the network to focus more on the hard pixels near the boundary by assigning different weights to different positions. Compared with 12 counterparts, experimental results on five benchmark datasets show that our method has better generalization and improves the state-of-the-art approach by 4.85% averagely in terms of the regional and boundary evaluation measures. In addition, our model is more efficient with an inference speed of 50.3 FPS when processing a 320 x 320 image. Code has been made available at https://github.com/lesonly/GFINet.

A multi-exit network is an important technique for achieving adaptive inference by dynamically allocating computational resources based on different input samples. The existing works mainly treat the final classifier as the teacher, enhancing the classification accuracy by transferring knowledge to the intermediate classifiers. However, this traditional self-distillation training strategy only utilizes the knowledge contained in the final classifier, neglecting potentially distinctive knowledge in the other classifiers. To address this limitation, we propose a novel multi-level collaborative self-distillation learning strategy (MLCSD) that extracts knowledge from all the classifiers. MLCSD dynamically determines the weight coefficients for each classifier’s contribution through a learning process, thus constructing more comprehensive and effective teachers tailored to each classifier. These new teachers transfer the knowledge back to each classifier through a distillation technique, thereby further improving the network’s inference efficiency. We conduct experiments on three datasets, CIFAR10, CIFAR100, and Tiny-ImageNet. Compared with the baseline network that employs traditional self-distillation, our MLCSD-Net based on ResNet18 enhances the average classification accuracy by 1.18%. The experimental results demonstrate that MLCSD-Net improves the inference efficiency of adaptive inference applications, such as anytime prediction and budgeted batch classification. Code is available at https://github.com/deepzlk/MLCSD-Net.

Objective. Sleep staging studies on single-channel EEG mainly exploit deep learning methods that combine convolutional neural networks (CNNs) and recurrent neural networks. However, when typical brain waves (such as K-complexes or sleep spindles) that identify sleep stages span two epochs, the abstract process of a CNN extracting features from each sleep stage may cause the loss of boundary context information. This study attempts to capture the boundary context, which contains the characteristics of brain waves during sleep stage transition, to improve the performance of sleep staging. Approach. In this paper we propose a fully convolutional network with boundary temporal context refinement, called BTCRSleep (Boundary Temporal Context Refinement Sleep). The boundary temporal context refinement module refines the boundary information on sleep stages on the basis of extracting multi-scale temporal dependences between epochs and enhances the abstract capability of the boundary temporal context. In addition, we design a class-aware data augmentation method to effectively learn the boundary temporal context between the minority class and other sleep stages. Main results. We evaluate the performance of our proposed network using four public datasets: the 2013 version of Sleep-EDF Expanded (SEDF), the 2018 version of Sleep-EDF Expanded (SEDFX), the Sleep Heart Health Study (SHHS) and CAP Sleep Database (CAP). The evaluation results on the four datasets showed that our model obtains the best total accuracy and kappa score compared with state-of-the-art methods. On average, accuracies of 84.9% in SEDF, 82.9% in SEDFX, 85.2% in SHHS and 76.9% in CAP are obtained under subject-independent cross-validation. We demonstrate that the boundary temporal context contributes to the improvement in capturing the temporal dependences across different epochs.

Current methods aggregate multi-level features or introduce edge and skeleton to get more refined saliency maps. However, little attention is paid to how to obtain the complete salient object in cluttered background, where the targets are usually similar in color and texture to the background. To handle this complex scene, we propose a sharp eyes network (SENet) that first seperates the object from scene, and then finely segments it, which is in line with human visual characteristics, i.e., to look first and then focus. Different from previous methods which directly integrate edge or skeleton to supplement the defects of objects, the proposed method aims to utilize the expanded objects to guide the network obtain complete prediction. Specifically, SENet mainly consists of target separation (TS) brach and object segmentation (OS) branch trained by minimizing a new hierarchical difference aware (HDA) loss. In the TS branch, we construct a fractal structure to produce saliency features with expanded boundary via the supervision of expanded ground truth, which can enlarge the detail difference between foreground and background. In the OS branch, we first aggregate multi-level features to adaptively select complementary components, and then feed the saliency features with expanded boundary into aggregated features to guide the network obtain complete prediction. Moreover, we propose the HDA loss to further improve the structural integrity and local details of the salient objects, which assigns weight to each pixel according to its distance from the boundary hierarchically. Hard pixels with similar appearance in border region will be given more attention hierarchically to emphasize their importance in completeness prediction. Comprehensive experimental results on five datasets demonstrate that the proposed approach outperforms the state-of-the-art methods both quantitatively and qualitatively.