Research map / Architectures and efficiency / Attention and transformers
ViTAEv2: Vision Transformer Advanced by Exploring Inductive Bias for Image Recognition and Beyond
Qiming Zhang, Yufei Xu, Jing Zhang, Dacheng Tao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add PCM to reduction cell (ablation) | ViTAE-T · ImageNet-1K (100 ep) | top-1 accuracy: 69.8 -> 71.7 | Improved |
| add pre fusion and BN in normal cell on top of PCM (ablation) | ViTAE-T · ImageNet-1K (100 ep) | top-1 accuracy: 71.7 -> 72.6 | Improved |
| add BN to pre fusion normal cell (ablation) | ViTAE-T · ImageNet-1K (100 ep) | top-1 accuracy: 69.1 -> 69.9 | Improved |
| raise dilation set to [1,2,3,4,5] (ablation) | ViTAE-T · ImageNet-1K (100 ep) | top-1 accuracy: 69.9 -> 68.9 | Hurt |
| use smaller dilation rates in deeper reduction cells (ablation) | ViTAE-T · ImageNet-1K (100 ep) | top-1 accuracy: 69.2 -> 69.8 | Improved |
| replace Performer attention with window attention in stage 1 (ablation) | ViTAEv2 · ImageNet-1K | top-1 accuracy: 82.2 -> 82.7 | Improved |
| use window attention in all four stages instead of first two (ablation) | ViTAEv2 · ImageNet-1K | top-1 accuracy: 82.6 -> 82.2 | Hurt |
| add window shifting and relative position encoding (ablation) | ViTAEv2-S · ImageNet-1K | top-1 accuracy: 82.6 -> 82.7 | No clear effect |
| add 1x1 convolution branch during pretraining (ablation) | ViTAE-L (MAE) · ImageNet-1K | top-1 accuracy: 84.1 -> 84.6 | Improved |
| add 3x3 convolution branch during pretraining (ablation) | ViTAE-L (MAE) · ImageNet-1K | top-1 accuracy: 84.1 -> 84.4 | Improved |
| replace Swin-T with ViTAEv2-S backbone | Swin-T · COCO Mask R-CNN 1x | box AP: 43.7 -> 46.3 | Improved |
| replace Swin-T with ViTAEv2-S backbone | Swin-T · ImageNet-1K | top-1 accuracy: 81.3 -> 82.6 | Improved |
Closest papers by meaning
- ViTAE: Vision Transformer Advanced by Exploring Intrinsic Inductive Bias (NeurIPS 2021 · similarity 0.95)
- Co-Advise: Cross Inductive Bias Distillation (CVPR 2022 · similarity 0.89)
- An Image is Worth More Than 16x16 Patches: Exploring Transformers on Individual Pixels (ICLR 2025 · similarity 0.88)
- Bridging the Gap Between Vision Transformers and Convolutional Neural Networks on Small Datasets (NeurIPS 2022 · similarity 0.88)
- AdaViT: Adaptive Vision Transformers for Efficient Image Recognition (CVPR 2022 · similarity 0.88)
- Linear Differential Vision Transformer: Learning Visual Contrasts via Pairwise Differentials (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
