search

berkeley-hipie / HIPIE

[NeurIPS2023] Code release for "Hierarchical Open-vocabulary Universal Image Segmentation"
https://people.eecs.berkeley.edu/~xdwang/projects/HIPIE/
MIT License
260 stars 19 forks source link

subpart inference? #11

Closed joshmyersdean closed 7 months ago

joshmyersdean commented 7 months ago

Hello!

Thank you for this great work! Is there any easy way to test subpart inference? Also is there a preferred way to fine-tune?

Thank you!

jacklishufan commented 7 months ago

https://github.com/berkeley-hipie/HIPIE/blob/1e3b657e158379b2ad391346bdc6070c666e8298/projects/HIPIE/hipie/demo_lib/part_segm_demo.py#L314 Him our predictor will generate both part and subparts. In particular output_refined is subparts and output_coarse is parts

joshmyersdean commented 7 months ago

thank you!