Sign in to view source links and access this dataset
Description
ActivityForensics is a benchmark dataset for localizing manipulated activity in videos, presented at CVPR 2026. The dataset was created by researchers including Peijun Bao, Anwei Luo, Gang Pan, Alex C. Kot, and Xudong Jiang. It was last updated on the platform in April 2026.
Use Cases
Training models to detect manipulated human activities in videos based on the benchmark's focus.
Benchmarking video localization algorithms for forensic applications.
Developing methods to distinguish between authentic and synthesized video content.
Strengths
Dataset is associated with a peer-reviewed CVPR 2026 conference paper, suggesting academic rigor.
The description explicitly states it is a comprehensive benchmark for a specific task.
Limitations
Column-level documentation is absent; field semantics must be inferred after download.
Row count is unknown, which may limit suitability assessment.
The dataset's license is unspecified, which could restrict usage.
Provenance
Source
Peijun Bao, Anwei Luo, Gang Pan, Alex C. Kot, Xudong Jiang
Collection Method
Likely created for research purposes as a benchmark for video forensics.
Time Range
null
Freshness
Last updated 2026-04-27 01:38:49; freshness should be verified.
Geography
null
License is unknown; users must verify terms before use.