Synthetic Data for training benthic species object detectors
Synthetic Data for training benthic species object detectors
3D viewer of synthetic image of an urchin barren taken from an Autonomous Underwater Vehicle (AUV) camera. The scene was generated with 3D modelling software Blender and based on the infinigen framework (Raistrick et al, 2023). Colour boards are included to demonstrate the absorption of light underwater.
Examples of urchin scenes. Segementation masks and depth maps are automatically generated with the synthetic images.
3D viewer of synthetic coral taken by an AUV.
References
References
Alexander Raistrick, Lahav Lipson, Zeyu Ma, Lingjie Mei, Mingzhe Wang, Yiming Zuo, Karhan Kayan, Hongyu Wen, Beining Han, Yihan Wang, Alejandro Newell, Hei Law, Ankit Goyal, Kaiyu Yang, & Jia Deng(2023). Infinite photorealistic worlds using procedural generation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (pp. 12630-12641).