ComfyUI Impact BIGdetector BBOX SEGS Released

By vramkickedin | | 2 min read
Image of text reads BIGdetector BBOX SEGS in a digtial style

Devourer56 has released ComfyUI Impact BIGdetector BBOX SEGS, a new custom node designed to implement a cascade object detection system within the ComfyUI environment. Developed over the course of half a day using AI assistance, this tool enables users to utilize YOLO BBOX/SEGM models for detailed object detection workflows.

The node functions as a magnifying glass for sequential detection on single images or aggregates multiple SEGS from a single source into one output with IoU filtering. It integrates directly with the established Impact Pack, outputting data in the standard SEGS format for compatibility with other nodes like SEGSPaste and SEGSDetailer.

Core features & technical capabilities

  • Sequential and parallel processing operation modes.
  • Multi-stage processing support.
  • Flexible configurations.
  • Integrated SEGS format output.
  • Filtering options.
  • Mask simplification using morphological operations.
  • Image bypass feature to return original images.

Developer comments

Devourer56 provides transparency regarding the origins of the code, stating,

'This was created using AI, and the node functionality would probably have been much simpler to implement using standard Impact Pack nodes, but I decided to go the extra mile and created a node in half a day using AI for my own needs.'

The developer acknowledges the nature of AI-generated code, noting that

'The guarantee of operation is the same as for any code written using AI.'

To ensure stability within user workflows, the documentation advises specific settings. The developer explains,

'It's recommended to use the node with Include masks in output = true. Without it, most nodes receiving SEGS will not work or will return errors.'

Grab ComfyUI Impact BIGdetector BBOX SEGS

Check out their project on their GitHub page here.