Please add the following papers

[mistake0316]

Text2Mesh : Their approach can modify a given mesh with given text/image information via CLIP text/image encoder.

Detecting Twenty-thousand Classes using Image-level Supervision : Which is a object detection research by facebook, they use CLIP text embedding as classifier weight.

The above papers are I want to add.

---

I think Crop-CLIP should be put at Object Detection

---

For more image manipuliation / generation applications, I summarized in my medium
You can add them if you think they are valueable.
Text-Driven Image Manipulation/Generation with CLIP

The list of above image is allocated at this google sheet

[yzhuoning]

Thanks for your input! I have included the above two papers. I will check your list later.