Closed-form continuous-time (CFC) neural networks have superior expressivity in modeling time series data compared with recurrent neural networks. CFC's lower training and inference overheads also make it appealin...
详细信息
Non-graphitized carbon(NGC)has been extensively utilized as carbonaceous anode in sodium-ion batteries(SIBs).However,more optimization to achieve competitive capacity and stability is still challenging for *** the stu...
详细信息
Non-graphitized carbon(NGC)has been extensively utilized as carbonaceous anode in sodium-ion batteries(SIBs).However,more optimization to achieve competitive capacity and stability is still challenging for *** the study,the dopant strategy is utilized to construct nitrogen/sulfur-doped non-graphitized carbon(N-NGC or S-NGC)shell decorated on three-dimensional graphene foam(GF)as a self-support *** highly disordered microstructures of heteroatom doped carbons are produced by applying a low-temperature pyrolysis treatment to precursors containing nitrogen and *** DFT calculations of Na-ion adsorption energies at diverse heteroatom sites show marginal-S,pyrrolic N and pyridinic N with more intensive Na-ion adsorption ability than middle-S,C=O and pristine *** N-NGC with dominant small graphitic regions delivers adsorption ability to Na-ion,while the S-NGC with significant single carbon lattice stripes demonstrates redox reaction with ***,in comparison with only adsorption-driven slope regions at high potential for N-NGC,the redox reaction-generated potentialplateau enables non-graphitized S-NGC superior discharge/charge capacity and cycle-stability in the slope *** work could provide deep insight into the rational design of non-graphitized carbon with rich microstructure and composition.
Semi-supervised learning has been an important approach to address challenges in extracting entities and relations from limited data. However, current semi-supervised works handle the two tasks (i.e., Named Entity Rec...
详细信息
Food volume prediction is critical for calculating calories in food analysis and nutritional evaluation. However, obtaining data and measuring volume from the 3D sensor camera concurrently is time-consuming. Additiona...
详细信息
Unlearnable examples (UEs) seek to maximize testing error by making subtle modifications to training examples that are correctly labeled. Defenses against these poisoning attacks can be categorized based on whether sp...
详细信息
Unlearnable examples (UEs) seek to maximize testing error by making subtle modifications to training examples that are correctly labeled. Defenses against these poisoning attacks can be categorized based on whether specific interventions are adopted during training. The first approach is training-time defense, such as adversarial training, which can mitigate poisoning effects but is computationally intensive. The other approach is pre-training purification, e.g., image short squeezing, which consists of several simple compressions but often encounters challenges in dealing with various UEs. Our work provides a novel disentanglement mechanism to build an efficient pre-training purification method. Firstly, we uncover rate-constrained variational autoencoders (VAEs), demonstrating a clear tendency to suppress the perturbations in UEs. We subsequently conduct a theoretical analysis for this phenomenon. Building upon these insights, we introduce a disentangle variational autoencoder (D-VAE), capable of disentangling the perturbations with learnable class-wise embeddings. Based on this network, a two-stage purification approach is naturally developed. The first stage focuses on roughly eliminating perturbations, while the second stage produces refined, poison-free results, ensuring effectiveness and robustness across various scenarios. Extensive experiments demonstrate the remarkable performance of our method across CIFAR-10, CIFAR-100, and a 100-class ImageNet-subset. Code is available at https://***/yuyi-sd/D-VAE. Copyright 2024 by the author(s)
This paper presents a modified finite control set model predictive control (FCS-MPC) that can significantly reduce computation time and maintain good quality of output voltage. The switching states of a three-phase fi...
详细信息
The explosive growth of social media means portrait editing and retouching are in high *** portraits are commonly captured and stored as raster images,editing raster images is non-trivial and requires the user to be h...
详细信息
The explosive growth of social media means portrait editing and retouching are in high *** portraits are commonly captured and stored as raster images,editing raster images is non-trivial and requires the user to be highly *** at developing intuitive and easy-to-use portrait editing tools,we propose a novel vectorization method that can automatically convert raster images into a 3-tier hierarchical *** base layer consists of a set of sparse diffusion curves(DCs)which characterize salient geometric features and low-frequency colors,providing a means for semantic color transfer and facial expression *** middle level encodes specular highlights and shadows as large,editable Poisson regions(PRs)and allows the user to directly adjust illumination by tuning the strength and changing the shapes of *** top level contains two types of pixel-sized PRs for high-frequency residuals and fine details such as pimples and *** train a deep generative model that can produce high-frequency residuals *** to the inherent meaning in vector primitives,editing portraits becomes easy and *** particular,our method supports color transfer,facial expression editing,highlight and shadow editing,and automatic *** quantitatively evaluate the results,we extend the commonly used FLIP metric(which measures color and feature differences between two images)to consider *** new metric,illumination-sensitive FLIP,can effectively capture salient changes in color transfer results,and is more consistent with human perception than FLIP and other quality measures for portrait *** evaluate our method on the FFHQR dataset and show it to be effective for common portrait editing tasks,such as retouching,light editing,color transfer,and expression editing.
This paper addresses the task of temporal activity localization (TAL). Although recent works have made significant progress in TAL research, almost all of them implicitly assume that the dense frame-level corresponden...
详细信息
This paper presents the finite control set model predictive control (FCS-MPC) with improved control objectives using optimized weighting factors. The Pareto curve obtained from the Particle Swarm optimizes the weighti...
详细信息
Current advances in deep learning have brought various breakthroughs in processing medical data. However, dealing with a limited number of medical datasets remains a challenge in deep learning and often leads to overf...
详细信息
暂无评论