AI Paper Atlas

Research map / Video / Video segmentation and tracking

TeD-SPAD: Temporal Distinctiveness for Self-Supervised Privacy-Preservation for Video Anomaly Detection

Joseph Fioresi, Ishan Rajendrakumar Dave, Mubarak Shah

ICCV 2023 · Video › Video segmentation and tracking · approach: Weakly supervised video anomaly detection · 35 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 temporally-distinct triplet loss to the anonymization utility branch (ablation)SPAct-style anonymizer · VISPR / UCF-Crimeprivacy cMAP; AUC: cMAP 52.71 -> 42.21; AUC 73.93 -> 74.81Improved
add temporal-invariance objective instead of distinctiveness (ablation)anonymization utility branch · UCF-CrimeAUC (%): 73.93 -> 69.52Hurt
use LLTC contrastive loss instead of triplet loss (ablation)temporal distinctiveness loss · UCF-Crime / VISPRAUC; cMAP: AUC 74.81 -> 75.06; cMAP 42.21 -> 42.44No clear effect
increase triplet loss weight from 0.1 to 1.0 (ablation)TeD-SPAD · UCF-Crime / VISPRAUC; cMAP: AUC 74.81 -> 70.56; cMAP 42.21 -> 55.26Hurt
lower margin from 1.0 to 0.5 (ablation)temporal triplet loss · UCF-CrimeAUC (%): 74.81 -> 62.12Hurt
raise margin from 1.0 to 2.0 (ablation)temporal triplet loss · UCF-CrimeAUC (%): 74.81 -> 67.30Hurt
use fixed larger negative clip distance 32 instead of random (ablation)temporal triplet loss · UCF-Crime / VISPRcMAP; AUC: cMAP 42.21 -> 28.69; AUC 74.81 -> 70.97Mixed
use fixed small negative clip distance 8 instead of random (ablation)temporal triplet loss · UCF-Crime / VISPRAUC; cMAP: AUC 74.81 -> 76.12; cMAP 42.21 -> 46.34Mixed
apply TeD-SPAD anonymizer instead of SPActprivacy-preserving anomaly detection · UCF-Crime / VISPRprivacy cMAP; AUC: cMAP 52.71 -> 42.21; AUC 73.93 -> 74.81Improved
anonymize XD-Violence videos (vs raw data)I3D + MGFN · XD-ViolenceAP (%): 73.72 -> 60.32Hurt
use standard Kinetics-pretrained I3D features (no anonymization)I3D features · VISPR feature-levelprivacy cMAP (%): 63.15 vs 52.30 (anonymized)Hurt

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