Efficient Vits On Imagenet 1K With Deit S

Metriken

GFLOPs

Top 1 Accuracy

Ergebnisse

Leistungsergebnisse verschiedener Modelle zu diesem Benchmark

Modellname	GFLOPs	Top 1 Accuracy	Paper Title	Repository
HVT-S-1	2.7	78.3	Scalable Vision Transformers with Hierarchical Pooling
DynamicViT (70%)	2.9	79.3	DynamicViT: Efficient Vision Transformers with Dynamic Token Sparsification
LTMP (45%)	2.3	78.6	Learned Thresholds Token Merging and Pruning for Vision Transformers
EViT (80%)	3.5	79.8	Not All Patches are What You Need: Expediting Vision Transformers via Token Reorganizations
BAT (70%)	3.0	79.6	Beyond Attentive Tokens: Incorporating Token Importance and Diversity for Efficient Vision Transformers
MCTF ($r=18$)	2.4	79.9	Multi-criteria Token Fusion with One-step-ahead Attention for Efficient Vision Transformers
BAT (60%)	2.6	79.3	Beyond Attentive Tokens: Incorporating Token Importance and Diversity for Efficient Vision Transformers
IA-RED$^2$	3.2	79.1	IA-RED$^2$: Interpretability-Aware Redundancy Reduction for Vision Transformers	-
AS-DeiT-S (50%)	2.3	78.7	Adaptive Sparse ViT: Towards Learnable Adaptive Token Pruning by Fully Exploiting Self-Attention	-
S$^2$ViTE	3.2	79.2	Chasing Sparsity in Vision Transformers: An End-to-End Exploration
BAT (20%)	1.6	76.4	Beyond Attentive Tokens: Incorporating Token Importance and Diversity for Efficient Vision Transformers
DPS-ViT	2.4	79.5	Patch Slimming for Efficient Vision Transformers	-
LTMP (80%)	3.8	79.8	Learned Thresholds Token Merging and Pruning for Vision Transformers
SPViT	3.3	78.3	Pruning Self-attentions into Convolutional Layers in Single Path
DiffRate	2.9	79.8	DiffRate : Differentiable Compression Rate for Efficient Vision Transformers
AS-DeiT-S (65%)	3.0	79.6	Adaptive Sparse ViT: Towards Learnable Adaptive Token Pruning by Fully Exploiting Self-Attention	-
Base (DeiT-S)	4.6	79.8	Training data-efficient image transformers & distillation through attention	-
EViT (90%)	4.0	79.8	Not All Patches are What You Need: Expediting Vision Transformers via Token Reorganizations
EViT (60%)	2.6	78.9	Not All Patches are What You Need: Expediting Vision Transformers via Token Reorganizations
PPT	2.9	79.8	PPT: Token Pruning and Pooling for Efficient Vision Transformers

0 of 41 row(s) selected.