Grounded-SAM colab notebook unable to detect objects of interest

IDEA-Research / Grounded-Segment-Anything

Grounded SAM: Marrying Grounding DINO with Segment Anything & Stable Diffusion & Recognize Anything - Automatically Detect , Segment and Generate Anything

https://arxiv.org/abs/2401.14159

Apache License 2.0

14.22k stars 1.31k forks source link

Grounded-SAM colab notebook unable to detect objects of interest #426

Open dinkarjuyal opened 6 months ago

dinkarjuyal commented 6 months ago

I am using the colab notebook automated-dataset-annotation-and-evaluation-with-grounding-dino-and-sam.ipynb provided in the repo to segment the following image - vision_test_pic I am passing the following classes to the inference code - ['person', 'mirror'] However the code is unable to detect these objects in the image provided. I have tried different values of box and text threshold, any suggestions on specific params to tune?