The dataset supports a broad range of tasks, including:
- Panoptic Segmentation
- Instance Segmentation
- Semantic Segmentation
- Object Detection
- Pretraining for Open-Vocabulary Segmentation
- Mask-Conditional Image Synthesis
As we advance into the Era of Big Data, the vast scale and versatility of datasets pave the way for the development of even more innovative tasks. Therefore, we invite the research community to explore its potential and contribute to groundbreaking advancements in the field.
Key Features:
- Extensive Image Collection: Approximately 383,000 high-quality images.
- Verified Annotations: 5.18 million meticulously verified panoptic segmentation masks.
- Versatile Applications: Suitable for a wide array of segmentation and detection tasks.
- Innovative Validation Set: COCONut-val provides a robust testbed for benchmarking.
Discover our dataset and open up new opportunities in segmentation and object detection research. Furthermore, join us in expanding the limits of what’s achievable with our detailed and adaptable data resources.