Depth Estimation 실행 결과 비교

같은 이미지를 기준으로 3개의 Depth Estimation 도출을 진행.

 

1. SegmentAnyRGBD

https://github.com/Jun-CEN/SegmentAnyRGBD

GitHub - Jun-CEN/SegmentAnyRGBD: Segment Any RGBD

 

2. Depth-Anything

https://github.com/LiheYoung/Depth-Anything

GitHub - LiheYoung/Depth-Anything: [CVPR 2024] Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data. Foundation Mo

 

3. Depth-Anything-V2

https://github.com/DepthAnything/Depth-Anything-V2

GitHub - DepthAnything/Depth-Anything-V2: [NeurIPS 2024] Depth Anything V2. A More Capable Foundation Model for Monocular Depth

 

가장 최신 방법론인 Depth-Anything-V2의 결과가 훨씬 더 멀리 인식 가능하며, 객체 간 Segment 또한 뛰어남을 확인함.