AI Paper Atlas

Research map / Architectures and efficiency / CNN design

Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation

Zhenda Xie, Zheng Zhang, Xizhou Zhu, Gao Huang, Stephen Lin

ECCV 2020 · Architectures and efficiency › CNN design · approach: Dense matching and estimation CNNs

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 changedWhereMeasuredEffect
use learnable RBF interpolation kernel instead of average pooling (ablation)Ours (ResNet-101) · COCO2017 val detectionGFLOPs: 110.0 -> 96.6Improved
use learnable RBF interpolation kernel instead of plain convolution (ablation)Ours (ResNet-101) · COCO2017 val detectionGFLOPs: 109.8 -> 96.6Improved
add interpolation instead of filling zeros (ablation)Ours (ResNet-101) · COCO2017 val detectionGFLOPs: 164.9 -> 96.6Improved
add interpolation instead of reusing features (ablation)Ours (ResNet-101) · COCO2017 val detectionGFLOPs: 226.1 -> 96.6Improved
add grid prior (stride 11) (ablation)Ours (ResNet-101) · COCO2017 val detectionmAP: 41.8 -> 42.0No clear effect
add grid prior at loss weight 0.2 (ablation)Ours (ResNet-101) · COCO2017 val detectionmAP: 32.63 -> 40.69Improved
add grid prior at loss weight 0.02 (ablation)Ours (ResNet-101) · COCO2017 val detectionmAP: 42.61 -> 43.28Improved
add grid prior to reduce test-time variance (ablation)Ours (ResNet-101) · COCO2017 val detectionmAP std over 5 tests: 0.09 -> 0.02Improved
use stochastic Gumbel-Softmax sampling instead of deterministic Gumbel-Softmax (ablation)Ours (ResNet-101) · COCO2017 val detectionGFLOPs at 42.0 mAP: 121.5 -> 96.6Improved
use stochastic sampling instead of ReLU gating (ablation)Ours (ResNet-101) · COCO2017 val detectionGFLOPs at ~42 mAP: 181.9 -> 96.6Improved
use sampling-interpolation instead of baseline without samplingResNet-101 detector · COCO2017 val detectionmAP / GFLOPs: 43.4 mAP 289.5 GFLOPs -> 43.3 mAP 160.4 GFLOPsImproved
use sampling-interpolation on CityscapesSegmentation baseline · Cityscapes val segmentationmean IoU / GFLOPs: 80.8 mean IoU 920.6 GFLOPs -> 80.6 mean IoU 373.2 GFLOPsImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy