赵智轩 commited on
Commit
74cf9f1
·
1 Parent(s): 3dd0dfd

Add top-level paper link to dataset card

Browse files
Files changed (1) hide show
  1. README.md +2 -0
README.md CHANGED
@@ -27,6 +27,8 @@ configs:
27
 
28
  # PerceptionComp
29
 
 
 
30
  PerceptionComp is a benchmark for complex perception-centric video reasoning. It focuses on questions that cannot be solved from a single frame, a short clip, or a shallow caption. Models must revisit visually complex videos, gather evidence across temporally separated segments, and combine multiple perceptual cues before answering.
31
 
32
  ## Dataset Details
 
27
 
28
  # PerceptionComp
29
 
30
+ Links: [Paper](https://arxiv.org/abs/2603.26653) | [GitHub](https://github.com/hrinnnn/PerceptionComp)
31
+
32
  PerceptionComp is a benchmark for complex perception-centric video reasoning. It focuses on questions that cannot be solved from a single frame, a short clip, or a shallow caption. Models must revisit visually complex videos, gather evidence across temporally separated segments, and combine multiple perceptual cues before answering.
33
 
34
  ## Dataset Details