Domain-Transferred Synthetic Data Generation for Improving Monocular Depth Estimation
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866916232304787456 |
|---|---|
| author | Lee, Seungyeop Peterson, Knut Arezoomandan, Solmaz Cai, Bill Li, Peihan Zhou, Lifeng Han, David |
| author_facet | Lee, Seungyeop Peterson, Knut Arezoomandan, Solmaz Cai, Bill Li, Peihan Zhou, Lifeng Han, David |
| contents | A major obstacle to the development of effective monocular depth estimation algorithms is the difficulty in obtaining high-quality depth data that corresponds to collected RGB images. Collecting this data is time-consuming and costly, and even data collected by modern sensors has limited range or resolution, and is subject to inconsistencies and noise. To combat this, we propose a method of data generation in simulation using 3D synthetic environments and CycleGAN domain transfer. We compare this method of data generation to the popular NYUDepth V2 dataset by training a depth estimation model based on the DenseDepth structure using different training sets of real and simulated data. We evaluate the performance of the models on newly collected images and LiDAR depth data from a Husky robot to verify the generalizability of the approach and show that GAN-transformed data can serve as an effective alternative to real-world data, particularly in depth estimation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_01113 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Domain-Transferred Synthetic Data Generation for Improving Monocular Depth Estimation Lee, Seungyeop Peterson, Knut Arezoomandan, Solmaz Cai, Bill Li, Peihan Zhou, Lifeng Han, David Computer Vision and Pattern Recognition Artificial Intelligence Image and Video Processing A major obstacle to the development of effective monocular depth estimation algorithms is the difficulty in obtaining high-quality depth data that corresponds to collected RGB images. Collecting this data is time-consuming and costly, and even data collected by modern sensors has limited range or resolution, and is subject to inconsistencies and noise. To combat this, we propose a method of data generation in simulation using 3D synthetic environments and CycleGAN domain transfer. We compare this method of data generation to the popular NYUDepth V2 dataset by training a depth estimation model based on the DenseDepth structure using different training sets of real and simulated data. We evaluate the performance of the models on newly collected images and LiDAR depth data from a Husky robot to verify the generalizability of the approach and show that GAN-transformed data can serve as an effective alternative to real-world data, particularly in depth estimation. |
| title | Domain-Transferred Synthetic Data Generation for Improving Monocular Depth Estimation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Image and Video Processing |
| url | https://arxiv.org/abs/2405.01113 |