AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Improved Handling of Motion Blur in Online Object Detection

Mohamed Sayed, Gabriel Brostow

CVPR 2021 · Detection and segmentation › Open-set and few-shot detection · approach: Open-set and unknown object detection · 46 citations (OpenAlex)

Check 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
expand bounding-box labels to cover the blur-kernel displacement (ablation)blur-augmented Faster R-CNN · COCO minival E5, P1 blurmAP@0.5: 31.37 -> 38.35Improved
specialize detector networks by exposure and select a network using a blur estimator (ablation)expanded-label blur-augmented Faster R-CNN · COCO minival E5, P1 blurmAP@0.5: 38.35 -> 43.78Improved
specialize networks by blur type instead of exposure (ablation)expanded-label blur-augmented Faster R-CNN · COCO minival E5, P1 blurmAP@0.5: 43.78 -> 40.81Hurt
center blur kernels to align blurred objects with detection labels (ablation)blur-augmented detector · COCO minival, motion blurmAP@0.5: up to 8-10mAP@50Improved
deblur the image before detection (ablation)Faster R-CNN · COCO minival E5, P1 blur, standard labelsmAP@0.5: 4.58 -> 6.26Improved
undersample the input along blur axes and reverse-scale backbone activations (ablation)Faster R-CNN · COCO minival E5, P1 blur, standard labelsmAP@0.5: 4.58 -> 31.39Improved
apply spatial AugMix without adapting the labels (ablation)blur-augmented Faster R-CNN · COCO minival E5, P1 blur, standard labelsmAP@0.5: 4.58 -> 8.21Improved
apply pixel-level AugMix (ablation)Faster R-CNN · COCO minival E5, P1 blur, standard labelsmAP@0.5: 4.58 -> 4.48No clear effect
adapt batch-normalization statistics from the test minibatch (ablation)Faster R-CNN · COCO minival E5, P1 blur, standard labelsmAP@0.5: 4.58 -> 4.40Hurt
train with uncentered blur kernels (ablation)blur-augmented Faster R-CNN · COCO minival E5, P1 blur, standard labelsmAP@0.5: 31.37 -> 25.66Hurt
combine blur augmentation with non-spatial AugMix (ablation)blur-augmented detector · RealBlur blurrymAP@0.5: 35.63 -> 36.82Improved
apply minibatch normalization to the expanded-label model (ablation)expanded-label blur-augmented Faster R-CNN · COCO minival E5, P1 blur, expanded labelsmAP@0.5: 38.35 -> 13.88Hurt

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