[KongchangAI]
Concept全景场景图生成 / Panoptic Scene Graph Generation

PSG

An extension of scene graph generation that represents objects using panoptic segmentation masks. It models all objects and their relations in an image as a graph structure and outputs a set of (subject, predicate, object) triples.

Timeline (last 90 days)

Nov 11

On the VG150 and PSG datasets, RaRe outperforms existing baselines in class-balanced recognition.

Unverified50%
Nov 6

Multimodal large language models (MLLMs) applied to the PSG task act as static observers, are limited to fixed-resolution views, often ignore key relations involving small or peripheral objects, and struggle to distinguish ambiguous predicates.

Unverified50%
Nov 6

PSG evaluation typically employs Recall@K (such as R@20 and R@50) as well as mR@K metrics for predicate categories, to measure the model's ability to recall fine-grained relations.

Unverified50%
Sep 6

The cost of a single PSG test in the United States can reach $1,000–$3,000.

Expired50%
Aug 27

实验室级别的多导睡眠图(PSG)通常同时采集多个EEG通道、眼电(EOG)、肌电(EMG)等多路生理信号,分期精度更高但设备成本高、佩戴复杂

Unverified50%

All Facts (4)

Source Articles