The official implementation of visual priors from the paper Mid-Level Visual Representations Improve Generalization and Sample Efficiency for Learning Visuomotor Policies. Arxiv preprint 2018.
Robust low-light image restoration and enhancement: classical, deep and adaptive hybrid pipelines.
cellprofiler-library implements CellProfiler's image processing and mathematical code, and is usable as a standalone library
Computer Vision and OCR library for detecting and analyzing UI elements
Evaluation for Detection and Tracking
Production-ready, backend-agnostic computer vision utilities for multi-object tracking and spatial reasoning
Cascaded motion detection, YOLO tracking and asynchronous VLM video inspection.
Specification (pydantic models + JSON Schema) for the vidlabel video annotation format
An open source image processing library
TorchSeg: Semantic Segmentation models for PyTorch
Visynx Python SDK: run your Visynx computer vision models (object detection, segmentation, pose, classification) hosted or on your own hardware, including NVIDIA Jetson.
Open, directory-based format and Python library for video annotations: time markers, boxes, masks, tracks.
Extracts the Machine Readable Zone (MRZ) data from document images
Cloud-native computer vision model training toolkit for Aegis AI
Train models with self-supervised learning in a single command
Segment objects in images from text, box or point prompts with Meta's SAM 3
Recover the data from figure images - lines, points, bars, actograms, heat maps - on your own computer: no uploads, no AI model.
cellprofiler-core implements the bulk of CellProfiler's non-gui functionality
Open-source, self-hosted image and video content moderation (nudity, violence, hate, weapons, drugs...) — a pip-installable alternative to cloud moderation APIs.
A computer vision library that supports Win, packaged with patches for visual libraries such as Opencv, matplotlib, and image. In the future, Pytorch will also be supported, all of which are personal (Auorui) engineering code experience.
Drop-in cv2 replacement (import cv2u as cv) with extra OpenCV utilities: video editing, data augmentation, COCO/YOLO annotations, ONNX inference, camera calibration, document scanning, image statistics, visualization, color clustering, Unicode paths and text, and more
D-FINE-seg: real-time object detection, keypoint estimation, instance and semantic segmentation framework
Efficient LoRA Fine-Tuning for Vision LLMs with advanced CLI and model zoo
A super-fast Python implementation of seam carving algorithm for intelligent image resizing.
Converts pixel-art-style images such as those from generative models or low-quality sprite web uploads to true resolution usable assets.
AI-agent-ready toolkit for 3D reverse engineering: image analysis, photogrammetry, mesh repair, primitive fitting and headless rendering.
A library for image computer vision
Chess image to FEN: recognize chess positions from screenshots, diagrams and photos of a chessboard. Finds the board, reads all 64 squares, detects orientation.
Rotation- and scale-invariant shape matching for machine vision — fast C++ core, simple Python API
NVIDIA Nemotron Page Elements v3: A specialized YOLOX-based object detection model for identifying tables, charts, infographics, titles, headers/footers, and text in document pages
AI-powered platform for extracting and analyzing data from technical drawings / CAD drawings to enhance manufacturing workflows.
A simple, high level, easy to use, open source computer vision library for python
Zero-dependency image processing library
Neural network backend for training and inference for animal pose estimation.
Easily computing clip embeddings and building a clip retrieval system with them
High-performance ML data loading: C++20 core with SIMD-accelerated transforms, pipe-operator composition, TBL v2 format, and a PyTorch-compatible DataLoader.
A powerful image annotation and segmentation tool with GUI, CLI, and Python API
Python package utils for computer vision
Understand unfamiliar ML notebooks with evidence-backed explanations inside Jupyter, without LLMs.
Anomaly detection for machine vision — train from good images only, structural and logical defects, with a dashboard
Image Annotation, including Histopathology slides; From H&E slide to training data in one conversation: nuclei, outlines and slide QC, exported to QuPath, COCO or YOLO. No GPU needed. Try it: https://agent.sachin-kale.com/#annotate
Extract tables from PDFs
Neo MCP server — bring Neo, the autonomous AI engineer for ML, LLM, GenAI & data workflows, into Claude Code, Cursor, Codex & other editors. Describe a task; Neo builds it and writes results to your repo.
detect and classify insects on a raspberry pi
Open-source, offline barcode detection, localization, and decoding for images and PDFs
Open-source AI toolkit for fashion tech and virtual try-on
Geodata toolbox