Codes and model parameters for our paper in AAAI-2022, "ACDNet: Adaptively Combined Dilated Convolution for Monocular Panorama Depth Estimation". (https://arxiv.org/abs/2112.14440)
Inference
Please download the pretrained model parameters here and put it in folder ./checkpoints. The inference process is executed with the following commands:
python infer.py --gpus 0 --checkpoints ./checkpoints/acdnet-m3d.pt --example ./examples/m3d.png
TODO
- models and parameters
- training codes