AI Paper Atlas

Research map / Video / Video-language understanding

Revisiting Uncertainty: On Evidential Learning for Partially Relevant Video Retrieval

Jun Li, Peifeng Lai, Xuhang Lou, Jinpeng Wang, Yuting Wang, Ke Chen, Yaowei Wang, Shu-Tao Xia

ICML 2026 · Video › Video-language understanding · approach: Text-video retrieval · 0 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 the clip-scale branch to frame-scale evidential learning (ablation)Holmes · TVRSumR: 193.1 -> 198.6Improved
add the frame-scale branch to clip-scale evidential learning (ablation)Holmes · TVRSumR: 194.1 -> 198.6Improved
add the aggregated co-evidential loss (ablation)Holmes · TVRSumR: 196.7 -> 198.6Improved
fuse both branches with UDPF instead of trusting frame branch (ablation)Holmes · TVRSumR: 194.7 -> 198.6Improved
fuse both branches with UDPF instead of trusting clip branch (ablation)Holmes · TVRSumR: 195.4 -> 198.6Improved
use UDPF fusion instead of Confidence Dominance (ablation)Holmes · TVRSumR: 197.0 -> 198.6Improved
add query label calibration (ablation)Holmes · TVRSumR: 196.3 -> 198.6Improved
calibrate polysemous queries beyond initial set S_n (ablation)Holmes · TVRSumR: 197.3 -> 198.6Improved
add intra-video evidential loss L_intra (ablation)Holmes · TVRSumR: 194.9 -> 198.6Improved
add intra-video evidential loss L_intra (ablation)Holmes · ActivityNet CaptionsSumR: 154.0 -> 156.8Improved
use flexible OT with dustbin instead of softmax assignment (ablation)Holmes · TVRSumR: 196.1 -> 198.6Improved
use flexible OT with dustbin instead of plain OT (ablation)Holmes · TVRSumR: 197.5 -> 198.6Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

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