Deeper ViT S 54

Deeper Vision Transformers (ViTs), such as ViT-S-54, aim to improve the accuracy and efficiency of image processing tasks by increasing the depth of the network architecture. Current research focuses on addressing training challenges associated with deeper ViTs, exploring novel training techniques like masked image residual learning, and optimizing models for specific hardware platforms and applications, including medical image analysis and efficient multi-task learning. These advancements are significant because they enhance the performance of ViTs while mitigating computational costs, leading to more practical and effective applications in various fields.

Papers

October 31, 2024

Reinforcement Learning Gradients as Vitamin for Online Finetuning Decision Transformers
Kai Yan, Alexander G. Schwing, Yu-Xiong Wang
Reinforcement Learning Offline Reinforcement Learning Decision Transformer Online Fine Tuning Deeper ViT S 54

October 10, 2024

M$^2$-ViT: Accelerating Hybrid Vision Transformers with Two-Level Mixed Quantization
Yanbiao Liang, Huihong Shi, Zhongfeng Wang
Vision Transformer Quantization Operator Hybrid CNN Transformer Mixed Quantization Deeper ViT S 54

September 11, 2024

DS-ViT: Dual-Stream Vision Transformer for Cross-Task Distillation in Alzheimer's Early Diagnosis
Ke Chen, Yifeng Wang, Yufei Zhou, Haohan Wang
Alzheimer'S Disease Classification Task Feature Representation Early Diagnosis Stage Pipeline Task Distillation Deeper ViT S 54

April 2, 2024

ViTamin: Designing Scalable Vision Models in the Vision-Language Era
Jieneng Chen, Qihang Yu, Xiaohui Shen, Alan Yuille, Liang-Chieh Chen
Vision Language Model Vision Model Vanilla Transformer Deeper ViT S 54 Scalable Vision

March 4, 2024

VITAMIN: A Compositional Framework for Model Checking of Multi-Agent Systems
Angelo Ferrando, Vadim Malvone
Multi Agent System Verification Task Formal Verification Model Checking Mathematical Logic Deeper ViT S 54

November 16, 2023

I&S-ViT: An Inclusive & Stable Method for Pushing the Limit of Post-Training ViTs Quantization
Yunshan Zhong, Jiawei Hu, Mengzhao Chen, Rongrong Ji
Vision Transformer Continuum Limit Vision Task Layer Wise Quantization Deeper ViT S 54 Vision Transformer Quantization Quantization Granularity

September 25, 2023

Masked Image Residual Learning for Scaling Deeper Vision Transformers
Guoxi Huang, Hongtao Fu, Adrian G. Bors
Supervised ImageNet Deep Transformer Deeper ViT S 54

February 9, 2023

IH-ViT: Vision Transformer-based Integrated Circuit Appear-ance Defect Detection
Xiaoibin Wang, Shuang Gao, Yuntao Zou, Jianlan Guo, Chu Wang
Vision Paper Industrial Defect Integrated Circuit Design Deeper ViT S 54

January 18, 2023

ViT-AE++: Improving Vision Transformer Autoencoder for Self-supervised Medical Image Representations
Chinmay Prabhakar, Hongwei Bran Li, Jiancheng Yang, Suprosana Shit, Benedikt Wiestler, Bjoern Menze
Self Supervised Learning Transformer Based Supervised Autoencoder Contrastive Loss ViT Lens Self Supervised Medical Deeper ViT S 54

December 12, 2022

CLIP Itself is a Strong Fine-tuner: Achieving 85.7% and 88.0% Top-1 Accuracy with ViT-B and ViT-L on ImageNet
Xiaoyi Dong, Jianmin Bao, Ting Zhang, Dongdong Chen, Shuyang Gu, Weiming Zhang, Lu Yuan, Dong Chen, Fang Wen, Nenghai Yu
Zero Shot Fine Tuning Supervised ImageNet ViT Lens Deeper ViT S 54

October 26, 2022

M$^3$ViT: Mixture-of-Experts Vision Transformer for Efficient Multi-task Learning with Model-Accelerator Co-design
Hanxue Liang, Zhiwen Fan, Rishov Sarkar, Ziyu Jiang, Tianlong Chen, Kai Zou, Yu Cheng, Cong Hao, Zhangyang Wang
Multi Task Learning Mixture of Expert Co Design Single Task Deeper ViT S 54

May 19, 2022

TRT-ViT: TensorRT-oriented Vision Transformer
Xin Xia, Jiashi Li, Jie Wu, Xing Wang, Xuefeng Xiao, Min Zheng, Rui Wang
Vision Transformer Transformer Based High Performing 1D ConvNet Tensor Compiler Deeper ViT S 54

May 12, 2022

Vision Transformer: Vit and its Derivatives
Zujun Fu
Convolutional Neural Network Vision Transformer Transformer Based Attention Based Encoder Decoder Derivative Process Deeper ViT S 54

Deeper ViT S 54

Papers

Reinforcement Learning Gradients as Vitamin for Online Finetuning Decision Transformers

M$^2$-ViT: Accelerating Hybrid Vision Transformers with Two-Level Mixed Quantization

DS-ViT: Dual-Stream Vision Transformer for Cross-Task Distillation in Alzheimer's Early Diagnosis

ViTamin: Designing Scalable Vision Models in the Vision-Language Era

VITAMIN: A Compositional Framework for Model Checking of Multi-Agent Systems

I&S-ViT: An Inclusive & Stable Method for Pushing the Limit of Post-Training ViTs Quantization

Masked Image Residual Learning for Scaling Deeper Vision Transformers

IH-ViT: Vision Transformer-based Integrated Circuit Appear-ance Defect Detection

ViT-AE++: Improving Vision Transformer Autoencoder for Self-supervised Medical Image Representations

CLIP Itself is a Strong Fine-tuner: Achieving 85.7% and 88.0% Top-1 Accuracy with ViT-B and ViT-L on ImageNet

M$^3$ViT: Mixture-of-Experts Vision Transformer for Efficient Multi-task Learning with Model-Accelerator Co-design

TRT-ViT: TensorRT-oriented Vision Transformer

Vision Transformer: Vit and its Derivatives