AI Paper Atlas

Research map / 3D vision / Human pose estimation

FAFA: Frequency-Aware Flow-Aided Self-Supervision for Underwater Object Pose Estimation

Jingyi Tang, Gu Wang, Zeyu Chen, Shengquan Li, Xiu Li, Xiangyang Ji

ECCV 2024 · 3D vision › Human pose estimation · approach: Instance-level 6D object pose · 4 citations (OpenAlex)

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
add frequency-aware FFT augmentation (amplitude mix and dropout) in synthetic pre-training (ablation)FAFA pre-training · DeepURLADD 0.1d: 50.18 -> 73.36Improved
add amplitude dropout during pre-training (ablation)FAFA pre-training · DeepURLADD 0.1d: 55.03 -> 73.36Improved
add amplitude mix during pre-training (ablation)FAFA pre-training · DeepURLADD 0.1d: 58.54 -> 73.36Improved
apply FFT augmentation also on real images in self-supervision (ablation)FAFA self-supervision · DeepURLMEAN: 73.46 -> 67.71Hurt
add photometric image-level loss L_photo (ablation)FAFA self-supervision · DeepURL5 deg: 50.31 -> 62.02Improved
add warp-mask loss L_warp-mask (ablation)FAFA self-supervision · DeepURL5 cm: 43.90 -> 49.36Improved
add feature-level alignment on top of image-level losses (ablation)FAFA self-supervision · DeepURLMEAN: 70.68 -> 73.46Improved
add self-supervised training on unlabeled real imagesFAFA pre-trained · ROV6D Pool-OccludedMEAN: 85.06 -> 91.39Improved
use FAFA self-supervision instead of PFCPFC · ROV6D Pool-Occluded5 cm: 72.10 -> 86.71Improved
use FAFA instead of PFC on DeepURLPFC · DeepURLADD 0.1d: 98.87 -> 97.64Hurt
use FAFA instead of PFC on DeepURLPFC · DeepURL5 cm: 44.10 -> 50.92Improved
add feature-level loss alone without image-level losses (ablation)FAFA self-supervision · DeepURLADD 0.1d: 97.00 -> 95.73Hurt

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