atlas

Research map / Architectures and efficiency / Neural architecture search

LightTrack: Finding Lightweight Neural Networks for Object Tracking via One-Shot Architecture Search

CVPR 2021 · Architectures and efficiency › Neural architecture search · approach: Bi-level differentiable NAS · 254 citations (OpenAlex)

Check your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's title only (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
Search the backbone instead of using the handcrafted backbone (ablation)LightTrack mobile tracker · VOT-19EAO: 0.268 -> 0.292Helped
Search the output feature layer in addition to the backbone (ablation)LightTrack mobile tracker · VOT-19EAO: 0.292 -> 0.307Helped
Search the head instead of using the handcrafted head (ablation)LightTrack mobile tracker · VOT-19EAO: 0.268 -> 0.297Helped
Search backbone, output feature layer, and head together (ablation)LightTrack mobile tracker · VOT-19EAO: 0.268 -> 0.333Helped
Add ImageNet pre-training to the searched architecture (ablation)LightTrack · VOT-19EAO: 21.3 -> 31.2 -> 33.3Helped
Use LightTrack LargeB rather than DiMP-50Object tracking · GOT-10kAO: 0.611 -> 0.623Helped
Use LightTrack Mobile rather than DiMP-50Object tracking · TrackingNet testP (%): 68.7 -> 69.5Helped
Use LightTrack LargeB rather than DiMP-50Object tracking · TrackingNet testAUC (%): 74.0 -> 73.3Hurt
Use LightTrack Mobile rather than Ocean (offline)Object tracking · VOT-19FLOPs (G): 20.3 -> 0.53Helped

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