Back to evaluations
Public evaluation
image_forensics_eval
Tests AutoGen multi-agent system detection of spliced or altered blot image samples.
Evaluation type
task based
Challenge
Build an AutoGen Multi-Agent Image Forensics Inspector with DeepSeek R1
Difficulty
Advanced
Rigor
Unspecified
Evaluation overview
How the linked challenge is judged: tasks, benchmarks, and criteria count.
Tasks
1
Benchmarks
0
Criteria
0
Task templates
Inputs and expected outputs.
Task 1
image_forensics_eval
Audits figure image and metadata to identify manipulated image panels.
Input format
JSON containing base64 image data or file path and figure caption text.
Output format
JSON containing manipulation_detected (boolean), confidence_score (float), and flagged_regions (list).