Search Results - "Han, Yizeng"

Dynamic Neural Networks: A Survey by Han, Yizeng, Huang, Gao, Song, Shiji, Yang, Le, Wang, Honghui, Wang, Yulin

“…Dynamic neural network is an emerging research topic in deep learning. Compared to static models which have fixed computational graphs and parameters at the…”
Get full text

Resolution Adaptive Networks for Efficient Inference by Yang, Le, Han, Yizeng, Chen, Xi, Song, Shiji, Dai, Jifeng, Huang, Gao

“…Adaptive inference is an effective mechanism to achieve a dynamic tradeoff between accuracy and computational cost in deep networks. Existing works mainly…”
Get full text

Fine-grained Recognition with Learnable Semantic Data Augmentation by Pu, Yifan, Han, Yizeng, Wang, Yulin, Feng, Junlan, Deng, Chao, Huang, Gao

“…Fine-grained image recognition is a longstanding computer vision challenge that focuses on differentiating objects belonging to multiple subordinate categories…”
Get full text

Latency-Aware Unified Dynamic Networks for Efficient Image Recognition by Han, Yizeng, Liu, Zeyu, Yuan, Zhihang, Pu, Yifan, Wang, Chaofei, Song, Shiji, Huang, Gao

“…Dynamic networks have become a pivotal area of study in deep learning due to their ability to selectively activate computing units (such as layers or channels)…”
Get full text

EfficientTrain++: Generalized Curriculum Learning for Efficient Visual Backbone Training by Wang, Yulin, Yue, Yang, Lu, Rui, Han, Yizeng, Song, Shiji, Huang, Gao

“…The superior performance of modern computer vision backbones (e.g., vision Transformers learned on ImageNet-1 K/22 K) usually comes with a costly training…”
Get full text

Spatially Adaptive Feature Refinement for Efficient Inference by Han, Yizeng, Huang, Gao, Song, Shiji, Yang, Le, Zhang, Yitian, Jiang, Haojun

“…Spatial redundancy commonly exists in the learned representations of convolutional neural networks (CNNs), leading to unnecessary computation on…”
Get full text

Adaptive Focus for Efficient Video Recognition by Wang, Yulin, Chen, Zhaoxi, Jiang, Haojun, Song, Shiji, Han, Yizeng, Huang, Gao

“…In this paper, we explore the spatial redundancy in video recognition with the aim to improve the computational efficiency. It is observed that the most…”
Get full text

OStr-DARTS: Differentiable Neural Architecture Search Based on Operation Strength by Yang, Le, Zheng, Ziwei, Han, Yizeng, Song, Shiji, Huang, Gao, Li, Fan

“…Differentiable architecture search (DARTS) has emerged as a promising technique for effective neural architecture search, and it mainly contains two steps to…”
Get full text

Towards Learning Spatially Discriminative Feature Representations by Wang, Chaofei, Xiao, Jiayu, Han, Yizeng, Yang, Qisen, Song, Shiji, Huang, Gao

“…The backbone of traditional CNN classifier is generally considered as a feature extractor, followed by a linear layer which performs the classification. We…”
Get full text

SimPro: A Simple Probabilistic Framework Towards Realistic Long-Tailed Semi-Supervised Learning by Du, Chaoqun, Han, Yizeng, Huang, Gao

“…Recent advancements in semi-supervised learning have focused on a more realistic yet challenging task: addressing imbalances in labeled data while the class…”
Get full text

FLatten Transformer: Vision Transformer using Focused Linear Attention by Han, Dongchen, Pan, Xuran, Han, Yizeng, Song, Shiji, Huang, Gao

“…The quadratic computation complexity of self-attention has been a persistent challenge when applying Transformer models to vision tasks. Linear attention, on…”
Get full text

ENAT: Rethinking Spatial-temporal Interactions in Token-based Image Synthesis by Ni, Zanlin, Wang, Yulin, Zhou, Renping, Han, Yizeng, Guo, Jiayi, Liu, Zhiyuan, Yao, Yuan, Huang, Gao

“…Recently, token-based generation have demonstrated their effectiveness in image synthesis. As a representative example, non-autoregressive Transformers (NATs)…”
Get full text

DeeR-VLA: Dynamic Inference of Multimodal Large Language Models for Efficient Robot Execution by Yue, Yang, Wang, Yulin, Kang, Bingyi, Han, Yizeng, Wang, Shenzhi, Song, Shiji, Feng, Jiashi, Huang, Gao

“…MLLMs have demonstrated remarkable comprehension and reasoning capabilities with complex language and visual data. These advances have spurred the vision of…”
Get full text

Exploring contextual modeling with linear complexity for point cloud segmentation by Chng, Yong Xien, Qiu, Xuchong, Han, Yizeng, Pu, Yifan, Cao, Jiewei, Huang, Gao

“…Point cloud segmentation is an important topic in 3D understanding that has traditionally has been tackled using either the CNN or Transformer. Recently, Mamba…”
Get full text

Dynamic Diffusion Transformer by Zhao, Wangbo, Han, Yizeng, Tang, Jiasheng, Wang, Kai, Song, Yibing, Huang, Gao, Wang, Fan, You, Yang

“…Diffusion Transformer (DiT), an emerging diffusion model for image generation, has demonstrated superior performance but suffers from substantial computational…”
Get full text

Semantic Refocused Tuning for Open-Vocabulary Panoptic Segmentation by Chng, Yong Xien, Qiu, Xuchong, Han, Yizeng, Ding, Kai, Ding, Wan, Huang, Gao

“…Open-vocabulary panoptic segmentation is an emerging task aiming to accurately segment the image into semantically meaningful masks based on a set of texts…”
Get full text

OStr-DARTS: Differentiable Neural Architecture Search based on Operation Strength by Yang, Le, Zheng, Ziwei, Han, Yizeng, Song, Shiji, Huang, Gao, Li, Fan

“…Differentiable architecture search (DARTS) has emerged as a promising technique for effective neural architecture search, and it mainly contains two steps to…”
Get full text

UniTTA: Unified Benchmark and Versatile Framework Towards Realistic Test-Time Adaptation by Du, Chaoqun, Wang, Yulin, Guo, Jiayi, Han, Yizeng, Zhou, Jie, Huang, Gao

“…Test-Time Adaptation (TTA) aims to adapt pre-trained models to the target domain during testing. In reality, this adaptability can be influenced by multiple…”
Get full text

DyFADet: Dynamic Feature Aggregation for Temporal Action Detection by Yang, Le, Zheng, Ziwei, Han, Yizeng, Cheng, Hao, Song, Shiji, Huang, Gao, Li, Fan

“…Recent proposed neural network-based Temporal Action Detection (TAD) models are inherently limited to extracting the discriminative representations and…”
Get full text

Mask Grounding for Referring Image Segmentation by Chng, Yong Xien, Zheng, Henry, Han, Yizeng, Qiu, Xuchong, Huang, Gao

“…Referring Image Segmentation (RIS) is a challenging task that requires an algorithm to segment objects referred by free-form language expressions. Despite…”
Get full text