Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks73 from public data
- A Survey on Vision Transformer4,578
This paper reviews these vision transformer models by categorizing them in different tasks and analyzing their advantages and disadvantages, and takes a brief look at the self-attention mechanism in computer vision, as it is the base component in transformer.
- Pre-Trained Image Processing Transformer2,345
To maximally excavate the capability of transformer, the IPT model is presented to utilize the well-known ImageNet benchmark for generating a large amount of corrupted image pairs and the contrastive learning is introduced for well adapting to different image processing tasks.
- DAFL: Data-Free Learning of Student Networks453
A novel framework for training efficient deep neural networks by exploiting generative adversarial networks (GANs) is proposed, where the pre-trained teacher networks are regarded as a fixed discriminator and the generator is utilized for derivating training samples which can obtain the maximum response on the discriminator.
- Data-Free Learning of Student Networks453
A novel framework for training efficient deep neural networks by exploiting generative adversarial networks (GANs) is proposed, where the pre-trained teacher networks are regarded as a fixed discriminator and the generator is utilized for derivating training samples which can obtain the maximum response on the discriminator.
- AdderNet: Do We Really Need Multiplications in Deep Learning?240
This paper develops a special back-propagation approach for AdderNets by investigating the full-precision gradient, and proposes an adaptive learning rate strategy to enhance the training procedure of Ad DerNets according to the magnitude of each neuron's gradient.
- AdderSR: Towards Energy Efficient Image Super-Resolution100
This paper thoroughly analyzes the relationship between an adder operation and the identity mapping and insert shortcuts to enhance the performance of SR models using adder networks and develops a learnable power activation for adjusting the feature distribution and refining details.
- 97
- Co-Evolutionary Compression for Unpaired Image Translation93
A novel co-evolutionary approach for reducing their memory usage and FLOPs simultaneously and synergistically optimized for investigating the most important convolution filters iteratively is developed.
- Distilling Portable Generative Adversarial Networks for Image Translation89
Qualitative and quantitative analysis by conducting experiments on benchmark datasets demonstrate that the proposed method can learn portable generative models with strong performance.
- A Semi-Supervised Assessor of Neural Architectures75
This paper employs an auto-encoder to discover meaningful representations of neural architectures and introduces a graph convolutional neural network to predict the performance of architectures based on the learned representations and their relation modeled by the graph.
- 73
- 66
- 59
- Distilling Semantic Priors from SAM to Efficient Image Restoration Models52
This work proposes a general framework to distill SAM's semantic knowledge to boost exiting IR models without interfering with their inference process, and designs a semantic-guided relation (SGR) module for SPD, which ensures semantic feature representation space consistency to fully distill the priors.
- GenDet: Towards Good Generalizations for AI-Generated Image Detection51
This paper addresses the unseen-generator detection problem by considering this task from the perspective of anomaly detection and proposes an adversarial teacher-student discrepancy-aware framework, which has achieved state-of-the-art on public benchmarks.
- CBQ: Cross-Block Quantization for Large Language Models44
CBQ incorporates a coarse-to-fine preprocessing strategy for suppressing weight and activation outliers, coupled with an adaptive LoRA-Rounding technique for precise weight quantization, enabling CBQ to not only handle extreme outliers effectively but also improve overall quantization accuracy.
- Brain-inspired Multilayer Perceptron with Spiking Neurons43
This paper incorporates the machanism of LIF neurons into the MLP models, to achieve better accuracy without extra FLOPs, and proposes a full-precision LIF operation to communicate be-tween patches, including horizontal LIF and vertical LIF in different directions.
- 40
- LegoNet: Efficient Convolutional Neural Networks with Lego Filters39
A split-transform-merge strategy for an efficient convolution by exploiting intermediate Lego feature maps is developed and Inspired by network engineering, it is suggested that an ordinary filter in the neural network can be upgraded to a sophisticated module as well.
- 37
- Vision Superalignment: Weak-to-Strong Generalization for Vision Foundation Models33
This paper introduces a novel and adaptively adjustable loss function for weak-to-strong supervision, which involves using a weaker model to supervise a stronger one, aiming to enhance the latter's capabilities beyond the former's limits.
- 33
- Species evolution and quality evaluation of four Coptis herbal medicinal materials in Southwest China26
The phylogenetic analysis of four Coptis herbal medicinal materials based on internal transcribed spacer sequences showed that C. chinensis had been considered to be the best herb, not only the high content in single constituent but also in total alkaloids.
- 22
- Transcriptome analyses provide insights into the difference of alkaloids biosynthesis in the Chinese goldthread (Coptis chinensis Franch.) from different biotopes22
Differential analyses of the transcriptome identified four pivotal candidate enzymes, including aspartate aminotransferaseprotein, polyphenol oxidase, primary-amine oxidationase, and tyrosine decarboxylase, were significantly differentially expressed and may be responsible for the difference of alkaloids contents in the accessions from different biotopes.
- DiC: Rethinking Conv3x3 Designs in Diffusion Models21
This work rethink one of the simplest yet fastest module in deep learning, 3x3 Convolution, to construct a scaled-up purely convolutional diffusion model, which surpasses existing diffusion transformers by considerable margins in terms of performance while keeping a good speed advantage.
- Pangu Pro MoE: Mixture of Grouped Experts for Efficient Sparsity20
Mixture of Grouped Experts (MoGE), which groups the experts during selection and balances the expert workload better than MoE in nature, is introduced, which leads to better expert load balancing and more efficient execution for both model training and inference on Ascend NPUs.
- 18
- Comparative Analysis of Genetic and Chemical Differences between Four Berberis Herbs Based on Molecular Phylogenetic and HPLC Methods16
The results showed that the discriminatory power of ITS, ITS2 and psbA-trnH was low, but the rbcL marker was highly effective and reliable for the species differentiation and the four Berberis species can be successfully classified based on phylogenetic analysis of the rBCL sequences.
- Instruct-IPT: All-in-One Image Processing Transformer via Weight Modulation15
This work proposes Instruct-IPT -- an All-in-One Image Processing Transformer that could effectively address manifold image restoration tasks with large inter-task gaps, and proposes weight modulation that adapts weights to specific tasks.
- 14
- The potential of H5N1 viruses to adapt to bovine cells varies throughout evolution13
It is demonstrated that virus adaptation to bovine cells is: (i) highly variable across 2.3.4.4.4b genotypes, (ii) limited in viruses predating the global expansion of this clade, and (iii) determined by the viral internal gene cassette, and not restricted to udder epithelial cells.
- 13
- LIPT: Latency-Aware Image Processing Transformer10
This paper proposes a novel non-volatile sparse masking self-attention that utilizes a pre-computing sparse mask to capture contextual information from a larger window with no extra computation overload, and devise the low-latency proportion LIPT block that substitutes memory-intensive operators with the combination of self-attention and convolutions to achieve practical speedup.
- IPT-V2: Efficient Image Processing Transformer using Hierarchical Attentions10
This work presents an efficient image processing transformer architecture with hierarchical attentions, called IPTV2, adopting a focal context self-attention (FCSA) and a global grid self-attention (GGSA) to obtain adequate token interactions in local and global receptive fields.
- Knowledge Distillation with Multi-granularity Mixture of Priors for Image Super-Resolution10
MiPKD, a multi-granularity mixture of prior KD framework is presented to facilitate efficient SR model through the feature mixture in a unified latent space and stochastic network block mixture to demonstrate the effectiveness of the proposed MiPKD method.
- Data Upcycling Knowledge Distillation for Image Super-Resolution9
The Data Upcycling Knowledge Distillation (DUKD) is presented, to transfer the teacher model's knowledge to the student model through the upcycled in-domain data derived from training data to improve the student model's performance and robustness.
- 8
- Towards Efficient Agents: A Co-Design of Inference Architecture and System8
Experiments demonstrate that through the synergistic collaboration of these methods, AgentInfer reduces ineffective token consumption by over 50%, achieving an overall 1.8-2.5 times speedup with preserved accuracy, underscore that optimizing for agentic task completion-rather than merely per-token throughput-is the key to building scalable, efficient, and self-improving intelligent systems.
- 7
- 6
- RNAS: Architecture Ranking for Powerful Networks.6
This work proposes to learn a performance predictor for ranking different models in the searching period using few networks pre-trained on the entire dataset and represents each neural architecture as a feature tensor and uses the predictor to further refining the representations of networks in the search space.
- Dynamic Frequency-Adaptive Knowledge Distillation for Speech Enhancement5
A novel dynamic frequency-adaptive knowledge distillation approach to effectively compress SE models, which dynamically assesses the model’s output, distinguishing between high and low-frequency components, and adapts the learning objectives to meet the unique requirements of different frequency bands.
- 5
- 5
- Collaboration of Teachers for Semi-supervised Object Detection4
The Collaboration of Teachers Framework (CTF) is proposed, which consists of multiple pairs of teacher and student models for training and greatly improves the utilization of unlabeled data and prevents the positive feedback cycle of unreliable pseudo-labels.
- IFT: Image Fusion Transformer for Ghost-free High Dynamic Range Imaging4
A novel image fusion transformer is proposed, referred to as IFT, which presents a fast global patch searching (FGPS) module followed by a self-cross fusion module (SCF) for ghost-free HDR imaging and achieves state-of-the-art performance both quantitatively and qualitatively.
- Crafting Efficient Neural Graph of Large Entropy4
This work proposes to use graph entropy as the measurement, which shows useful properties to craft high-quality neural graphs and enables the proposed efficient algorithm to construct them as the initial network architecture.
- Pangu Light: Weight Re-Initialization for Pruning and Accelerating LLMs3
Pangu Light is introduced, a framework for LLM acceleration centered around structured pruning coupled with novel weight re-initialization techniques designed to address this ``missing piece''.
- 3
- 2
- Model quantization for computing-in-memory: a survey2
A comprehensive review of model quantization methods for CIM-based accelerators from three perspectives: fixed precision quantization, mixed precision quantization, and optimization of quantized models is presented.
- Unshackling Context Length: An Efficient Selective Attention Approach through Query-Key Compression2
Efficient Selective Attention (ESA), a novel approach that extends context length by efficiently selecting the most critical tokens at the token level to compute attention, reduces the computational complexity of token selection by compressing query and key vectors into lower-dimensional representations.
- Transferable text data distillation by trajectory matching1
This work proposed a method that involves learning pseudo prompt data based on trajectory matching and finding its nearest neighbor ID to achieve cross-architecture transfer to achieve cross-architecture transfer and is the first data distillation work suitable for text generation tasks such as instruction tuning.
- Rethinking 1-bit Optimization Leveraging Pre-trained Large Language Models1
This paper introduces a consistent progressive training for both forward and backward, smoothly converting the full-precision weights into the binarized ones, and incorporates binary-aware initialization and dual-scaling compensation to reduce the difficulty of progressive training and improve the performance.
- 1
- 1
- –
- –
- –
- –
- Autoregressive Image Generation with Vision Full-view Prompt–
Inspired by CoT from the field of NLP, IGTR adds reasoning prompts for AR image generation to simulate the human reasoning process, which enhances contextual reasoning by allowing the model to first perceive overall distribution information before generating the image, and improve generation stability by increasing the inference steps.
- –
- –
- –
- –
- –
- –
- –
- –
- –
- Improving the Instruction Quality of Accounting and Statistics Courses in Correspondence Education–
The paper analyzes the problems existed in the courses and comes up with the solution to the problems.
- –
Publication data from OpenAlex; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-14. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.