AI Paper Atlas

Research map / Video / Video segmentation and tracking

An Egocentric Look at Video Photographer Identity

Yedid Hoshen, Shmuel Peleg

CVPR 2016 · Video › Video segmentation and tracking · approach: Layered video decomposition and dynamics · 39 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
Replace the hand-designed LPC descriptor and SVM with an end-to-end CNNPhotographer identification on EVPR · EVPR, same-day and one-week-later recognition, 12sTop-1 accuracy: 81% -> 90%; 76% -> 91%Improved
Combine predictions across longer video using MAPCNN photographer identification · FPSI, 50sidentification accuracy: 76% -> around 91%Improved
Use MAP aggregation rather than mode aggregationPhotographer identification over multiple short sequences · FPSIidentification accuracyImproved
Increase the video duration from 4s to 12sCNN photographer verification · EVPREER: 11.3% -> 8.1%Improved
Subtract each frame's mean optical-flow vector as simple stabilization (ablation)CNN and LPC identification · EVPR, same-day identificationaccuracy: CNN: 77% -> 71% (4s), 90% -> 86% (12s); LPC: 65% -> 59% (4s), 81% -> 72% (12s)Hurt
Use the learned CNN descriptor instead of LPCPhotographer verification · EVPR, 4s and 12sEER: 13.6% -> 11.3%; 9.6% -> 8.1%Improved
Scale optical-flow values by dividing by the square root of their absolute value (ablation)CNN optical-flow input · Photographer recognitionrecognition performanceImproved
Use a temporal convolutional first layer rather than spatially invariant convolutionCNN for egocentric photographer recognition · Photographer identificationidentification accuracyMixed

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