[KongchangAI]
Unverified50% confidenceFactExact time

The GRAC workflow has two steps: given an input scene graph, it first retrieves relevant "graph-description" paired samples from the training data; then it compares the retrieved samples with the instrument-tissue interactions in the input scene, uses contrastive features to adjust the retrieved description representations, and finally generates a textual description for the current scene.

1
Sources
50%
Confidence
Long-term
Relevance
11/9/2026
First Seen

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1150306
API
curl https://kongchang.com/api/v1/knowledge/claims/1150306
MCP
get_claim(id=1150306)