search

jiaqixuac / WResVLM

Towards Real-World Adverse Weather Image Restoration: Enhancing Clearness and Semantics with Vision-Language Models (ECCV 2024)
MIT License
12 stars 0 forks source link

Question about learnable weather prompts #2

Open TanMIMI opened 1 week ago

TanMIMI commented 1 week ago

Great work! I have a question regarding the weather prompts training. The learnable parameter is a float, but the CLIP text encoder requires text input. Could you explain how you set up the prompt and input it into the CLIP text encoder for training?

jiaqixuac commented 1 week ago

Hello, thank you for your interest. The input for CLIP can be embedding features. Please refer to https://github.com/KaiyangZhou/CoOp

TanMIMI commented 1 week ago

Thank you so much for your reply!

---- Replied Message ---- | From | Jiaqi @.> | | Date | 11/14/2024 14:12 | | To | @.> | | Cc | @.>@.> | | Subject | Re: [jiaqixuac/WResVLM] Question about learnable weather prompts (Issue #2) |

Hello, thank you for your interest. The input for CLIP can be embedding features. Please refer to https://github.com/KaiyangZhou/CoOp

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>