Go to OpenArt main site
Upload workflow

2D to 3D

5.0

0 reviews
73
10.2K
3.0K
9
Description

This is a workflow to compare prompt word inference effects, comparing the image recognition capabilities of gemini, clipinterrogator and image2prompt. By the way, I also compared the depth maps of miragold, depth anything and zoe depth.


https://github.com/ZHO-ZHO-ZHO/ComfyUI-Gemini

free gemini api key: https://makersuite.google.com/app/apikey

Add your Gemini_API_Key to the custom_nodes/ComfyUI-Gemini/config.json file.


https://github.com/zhongpei/Comfyui_image2prompt


Discussion

(No comments yet)

Loading...

Author

213
521.0K
9.2K
1.8M

No reviews yet

  • - latest (a year ago)

Primitive Nodes (8)

Anything Everywhere (6)

ClipInterrogator (1)

Image scale to side (1)

Custom Nodes (52)

ComfyUI

  • - LoadImage (1)

  • - CLIPTextEncode (7)

  • - PreviewImage (6)

  • - ControlNetLoader (6)

  • - ControlNetApply (6)

  • - ConditioningConcat (3)

  • - VAEDecode (3)

  • - CheckpointLoaderSimple (1)

  • - KSampler (3)

  • - SaveImage (3)

  • - VAEEncode (1)

ComfyUI Nodes for Inference.Core

  • - Scribble_XDoG_Preprocessor (3)

  • - Zoe_DepthAnythingPreprocessor (1)

  • - Zoe-DepthMapPreprocessor (1)

  • - Image2Text (1)

  • - LoadImage2TextModel (1)

  • - DisplayText_Zho (3)

  • - Gemini_API_S_Zho (1)

  • - MarigoldDepthEstimation (1)

Checkpoints (1)

samaritan3dCartoon_v40SDXL.safetensors

LoRAs (0)