search

Iven-Wu / CASA

code for Category-agnostic Skeletal Animal Reconstruction
Apache License 2.0
35 stars 2 forks source link

.npz discrepancy test_animal #3

Open BastiaanGroot opened 4 weeks ago

BastiaanGroot commented 4 weeks ago

Hi,

I have an issue when running python main.py -c config/real_world.yaml After running the clip_retrieve.py, the retrieval animal is correctly set. However, after constructing the data_utils.Optimization_data() object, the .npz and .png files are loaded for the test_animal. This is rather strange as we have as input of the test_animal the Annotations and JPEGImages from DAVIS, and thus no .npz files.

I'd hope to recieve clarification regarding this.

Thanks

Iven-Wu commented 4 weeks ago

Thank you for your questions.

Right now we only provide the code for synthetic animals. So the code loads .npz files from the dataset and the .npz file contains information such as optical flow, segmentation, etc..

For the DAVIS dataset or any other real-world datasets, you will need to use other models to calculate all those information.

And for real-world cases, only the PNG and mesh shapes are used during the process, image for retrieval and mesh for init shape.

BastiaanGroot commented 2 weeks ago

Thank you for the quick response, I have been working on replicating the results proposed in the paper using the DAVIS dataset. Can I maybe contact you in private for some questions regarding this?

Iven-Wu commented 2 weeks ago

Sure. You can find my personal email on my homepage.