You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Model Card for `rebotnix/kineva_silver_segmentation_person`

Person Segmentation Model – Trained by KINEVA, Built by REBOTNIX, Germany Current State: in production and re-training.

rb_person is a high-performance segmentation model trained specifically for person detection and segmentation. With a single class (person), it is optimized for accurate pixel-level segmentation of people across varied lighting, scale, pose, and background conditions. This model is well suited for applications such as people counting, pedestrian detection, background removal, and privacy-aware video processing.

Developed and maintained by REBOTNIX, Germany, https://rebotnix.com

About KINEVA

KINEVA is an automated training platform based on the MCP Agent system. It regularly delivers new visual computing models, all developed entirely from scratch. This approach enables the creation of customized models tailored to specific client requirements, which can be retrained and re-released as needed. The platform is particularly suited for applications that demand flexibility, adaptability, and technological precision—such as industrial image processing, smart city analytics, or automated object detection.

KINEVA is continuously evolving to meet the growing demands in the fields of artificial intelligence and machine vision. https://rebotnix.com/en/kineva

Example Predictions

Input Image	Detection Result





(More example visualizations coming soon)

Model Details

Architecture: KINEVA SILVER SEGMENTATION (custom training head with optimized anchor boxes)
Task: Person Segmentation (1 class: person)
Trained on: Custom person segmentation dataset with diverse scenes and conditions
Format: PyTorch .pth + ONNX and trt export available on request
Parameters: Silver Version
Training Framework: PyTorch + KINEVA + custom augmentation

Dataset

The model was trained on a curated person segmentation dataset, which includes:

1 class: person
Diverse backgrounds and lighting conditions
Varying poses, scales, and levels of occlusion
Indoor and outdoor scenes with single and multiple people

Intended Use

Intended Use	Not Intended Use
Person segmentation in images and video	Surveillance without human review
People counting and crowd analysis	Military / lethal applications
Background removal and replacement	Real-time tracking of individuals in critical situations
Pedestrian detection for automation	Unauthorized biometric identification

Limitations

May yield false positives in scenes with person-like shapes (mannequins, statues)
Not fine-tuned for thermal or night vision
Heavily occluded or overlapping people in crowded scenes may reduce segmentation accuracy
Very small or distant people may be missed depending on input resolution

Usage Example


from kineva import KINEVA_SEG

#initialize model
model = KINEVA_SEG(model="models/kineva_silver_segmentation_person.pth")

#run inference on image
final_boxes, final_scores, final_labels = model.detect("example_person1.jpg", threshold=0.35)

#draw detection
model.draw(final_boxes, final_scores, final_labels, output_path="./outputs/output_1.jpg")

Contact

For commercial use or re-training this model support, or dataset access, contact:

REBOTNIX
Email: communicate@rebotnix.com
Website: https://rebotnix.com

License

This model is released under CC-BY-NC-SA unless otherwise noted. For commercial licensing, please reach out to the contact email.

Downloads last month: -; Downloads are not tracked for this model. How to track

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support