Ikomia HUB

Train DEIMv2: Real-Time Object Detection Meets DINOv3

Infer DEIMv2: Real-Time Object Detection Meets DINOv3

DeepSeek-OCR document OCR to Markdown

Transform documents into structured markdown with intelligent content recognition and semantic tagging

Convert PDF to image

Run vision-language model series based on Qwen3

Run vision-language model series based on Qwen2.5

Train RF-DETR models

Inference with RF-DETR models

Inference with YOLOv11 pose estimation models

Train YOLO11 classification models.

Inference with YOLOv11 image classification models

Train YOLOv11 instance segmentation models.

Inference with YOLOv11 segmentation models

Train YOLOv11 object detection models.

Inference with YOLOv11 models

OCR-free model for document understanding

Remove background with rembg library

Inference for Segment Anything Model 2 (SAM2).

Depth Anything V2 is a highly practical solution for robust monocular depth estimation

Train YOLOv9 instance segmentation models.

Instance segmentation with YOLOv9 models

Run OWLv2 a zero-shot text-conditioned object detection model

Train YOLOv9 models

Object detection with YOLOv9 models

Depth Anything is a highly practical solution for robust monocular depth estimation

Algorithm to colorize grayscale image

Image restoration algorithms with Swin Transformer

Infer ByteTrack for object tracking

Train YOLOv8 classification models.