Research map / Detection and segmentation / CNN features
Parameter-Inverted Image Pyramid Networks for Visual Perception and Multimodal Understanding
Zhaokai Wang, Xizhou Zhu, Xue Yang, Gen Luo, Hao Li, Changyao Tian, Wenhan Dou, Junqi Ge, Lewei Lu, Yu Qiao, Jifeng Dai
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 |
|---|---|---|---|
| replace ViTDet-B by parameter-inverted pyramid PIIP-TSB at similar FLOPs | ViTDet-B · COCO Mask R-CNN 1x | box AP: 43.8 -> 46.6 | Improved |
| replace ViTDet-B by PIIP-TSB at lower FLOPs (243G vs. 463G) | ViTDet-B · COCO Mask R-CNN 1x | box AP / FLOPs: 43.8 -> 43.9 with 463G -> 243G | Improved |
| replace ViTDet-L by PIIP-SBL at similar FLOPs | ViTDet-L · COCO Mask R-CNN 1x | box AP: 46.8 -> 49.4 | Improved |
| replace InternViT-6B by PIIP-LH6B | InternViT-6B · COCO Mask R-CNN 1x | box AP: 53.8 -> 55.7 | Improved |
| replace InternViT-6B by PIIP-LH6B | InternViT-6B · ADE20K UperNet | mIoU: 58.36 -> 59.65 | Improved |
| use image pyramid via multi-scale training on a single branch (ablation) | ViTDet-B · COCO Mask R-CNN 1x | box AP: 43.8 -> 44.8 | Improved |
| use parameter-inverted resolution order instead of parameter-direct (ablation) | PIIP-TSB · COCO Mask R-CNN 1x | box AP: 42.6 -> 46.6 | Improved |
| use ConvNeXt for the high-resolution branch and ViT for lower ones instead of all ViT (ablation) | PIIP-TSB · COCO Mask R-CNN 1x | box AP: 44.3 -> 47.1 | Improved |
| merge outputs of all three branches instead of only branch B (ablation) | PIIP-TSB 1568/896/672 · COCO | box AP: 43.1 -> 46.6 | Improved |
| use deformable attention instead of regular attention (12 interactions) (ablation) | PIIP-TSB 1120/896/448 · COCO | box AP: 42.0 -> 43.9 | Improved |
| increase interactions from 0 to 12 with deformable attention (ablation) | PIIP-TSB 1120/896/448 · COCO | box AP: 41.3 -> 43.9 | Improved |
| add 2 interactions to the vision encoder (ablation) | PIIP-LLaVA CLIP-B+CLIP-L · multimodal 8-benchmark average | Avg: 64.4 -> 65.2 | Improved |
Closest papers by meaning
- Parameter-Inverted Image Pyramid Networks (NeurIPS 2024 · similarity 0.96)
- Feature Pyramid Networks for Object Detection (CVPR 2017 · similarity 0.88)
- Inverted Pyramid Multi-task Transformer for Dense Scene Understanding (ECCV 2022 · similarity 0.88)
- Feature Pyramid Transformer (ECCV 2020 · similarity 0.88)
- GraphFPN: Graph Feature Pyramid Network for Object Detection (ICCV 2021 · similarity 0.87)
- MMIU: Multimodal Multi-image Understanding for Evaluating Large Vision-Language Models (ICLR 2025 · similarity 0.84)
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.
