Skip to content

Dataset · chengyenhsieh

Occluded and invisible object annotation in video

chengyenhsieh/TAO-Amodal

TAO dataset extension featuring amodal bounding boxes and segmentation masks for partially or fully hidden objects.

What's inside

The dataset supplements the TAO dataset with amodal bounding boxes for objects concealed behind obstacles, completely invisible, or outside the frame boundaries. Modal segmentation masks are also included in the annotations. The card specifies fewer than 1,000 records with image and text modalities.

What you can do with it

  • Training detectors to track occluded objects in frames
  • Testing tracking and segmentation models in complex scenes
  • Developing computer vision systems for robotics
Tasks
Object detection
Size
up to 1K rows
Downloads this month
4.8K
Likes
31
Access
on request on Hugging Face

How to load it

from datasets import load_dataset

ds = load_dataset("chengyenhsieh/TAO-Amodal")