viperyl/comfyui-gconet-plus

ComfyUI-GCoNet-plus

Introduction

Group collaborative learning network, termed GCoNet+, which can effectively and efficiently co-salient objects in natural scenes. This repo pack GCoNet-plus as ComfyUI nodes, and make this fast SOTA model easier use for everyone.

In the first runtime, it needs 2.37 secodns for checkpoint loading and move to device. Then each image image required less time for inference, depends on you inference device.

Installation

1. Go to comfyUI custom_nodes folder, ComfyUI/custom_nodes/
2. git clone https://github.com/viperyl/ComfyUI-GCoNet-plus.git
3. pip install requirements.txt

Checkpoints Download

Make sure you have intall git lfs

1. go to comfyUI models folder, ComfyUI/models/
2. git clone https://huggingface.co/ViperYX/GCoNet_plus_checkpoints

Usage

The demo workflow placed in workflow/example_workflow.json

inference speed screenshot. batch mode can also speed up inference time, but i still have no idea how to present this function.

Others

This model is designed for fast detection, so the model inference speed is incrediable high. But I havn’t encoutnered any scenario that needs such high inference speed, so I only did very basic development on this repo. If you have any ideas or enhancments, please let me know, i am happy to do it. In the final, if this repo helped you, pleas give a star to the model author GCoNet_plus

comfyui-gconet-plus's Issues

an error message appears

[ERROR] An error occurred while retrieving information for the 'GCovNetPlus' node.
Traceback (most recent call last):
File "D:\AI\ComfyUI\server.py", line 414, in get_object_info
out[x] = node_info(x)
^^^^^^^^^^^^
File "D:\AI\ComfyUI\server.py", line 392, in node_info
info['input'] = obj_class.INPUT_TYPES()
^^^^^^^^^^^^^^^^^^^^^^^
File "D:\AI\ComfyUI\custom_nodes\ComfyUI-GCoNet-plus\GCovNet_plus_node.py", line 71, in INPUT_TYPES
"checkpoints": (checkpoints, {"default": checkpoints[0]}),
~~~~~~~~~~~^^^
IndexError: list index out of range

comfyui.log