Trainbot (Onlytrains) is a computer vision application that automatically detects trains passing a fixed camera, stitches the train's frames into a single composite image, and logs sightings to a SQLite database. It is designed to run on a Raspberry Pi with a camera (USB, Pi Cam 3, or video file) and can optionally upload results via FTP to a web server. The project is unmaintained as stated in the README. Architecturally, it is a monolith Go backend with three CLI binaries: `trainbot` (main detection loop), `confighelper` (interactive crop selection), and `cleanup` (orphaned blob removal). The backend captures frames using `pkg/vid` (abstraction over V4L2, libcamera-vid, ffmpeg), processes them through a stitch pipeline that detects motion, computes frame offsets via patch matching, fits a constant-acceleration model using RANSAC, and then blends frames into a single JPEG and an animated GIF. Results are stored in a SQLite database and served by a built-in HTTP server. The frontend is a Vue 3 static SPA that loads the SQLite database client-side to display train sightings, stats, and images with no server-side rendering. Key highlights include highly optimized image processing: the `pkg/pmatch` package offers four implementations of cosine-similarity patch search (slow Go, optimized Go, Cgo, Vulkan compute), and the `pkg/ransac` package provides a generic RANSAC fitter for the speed model. The codebase demonstrates strong engineering craftsmanship with a clear package structure, extensive tests, test data sets, and a CI pipeline. However, heavy reliance on Cgo and Vulkan, plus the explicit unmaintained status, introduce cross-compilation complexity and potential drift from modern tooling.
OrbbecSDK is a cross-platform C/C++ software development kit for Orbbec 3D cameras, supporting depth, color, infrared, and IMU sensors. It provides both a high-level Pipeline API for easy stream management and a low-level Device/Sensor API for fine-grained control. The architecture is library-based with a C API (ObSensor.h) and a C++ wrapper (ObSensor.hpp), following a consistent pattern of context creation, device enumeration, stream configuration, frame acquisition, and post-processing via composable filters. Key entry points include Context (device discovery), Pipeline (streaming), Device (property control), and Filter (alignment, point cloud generation, noise removal, etc.). Data flows involve configuring stream profiles, starting the pipeline, receiving frames (depth, color, IR, IMU), applying filters (e.g., depth-to-color alignment, HDR merge, spatial/temporal filtering), and optionally recording or replaying streams. External dependencies include OpenCV for example visualizations, cJSON for multi-device synchronization configuration, and platform-specific USB drivers. The SDK is used in robotics, AR/VR, industrial inspection, and 3D scanning applications.
VOID is a Netflix research codebase for removing objects from videos while also erasing their physical side effects, such as shadows, reflections, and secondary motion caused by the removed object. The core runtime is a two-pass diffusion inpainting system built on CogVideoX: pass 1 performs coarse object removal guided by a quadmask, and pass 2 refines temporal consistency using warped noise and optical-flow-inspired conditioning. The repo is primarily an application/research library rather than a pure data project: it ships a `videox_fun` Python package with model definitions, custom diffusion pipelines, inference and training scripts, a FastAPI/Ray inference API, Gradio UIs, and a Colab notebook, alongside synthetic data-generation code that uses Blender and Kubric.
Scal3R is a research codebase for CVPR 2026 Highlight paper on large-scale 3D reconstruction from unordered images using test-time training. The core contribution is a novel bidirectional attention mechanism with fast weight updates (TTT layers) that adapts during inference to improve reconstruction quality for large scenes. The repository provides an end-to-end inference pipeline: given a folder of images, it estimates camera poses, depth maps, and produces a colored point cloud. The architecture builds on a DINOv2 backbone and VGGT-based aggregator with custom attention heads, loop detection via SALAD, and pose graph optimization with submap management. The codebase is well-structured with separate modules for data loading, model definition, inference orchestration, and visualization tools. It includes fallback implementations for key components (e.g., Python-only loop solver, simple correlation-based loop detector) to improve portability, though GPU and CUDA are assumed. The repository is primarily a research artifact with clean utilities that could be extracted for other 3D vision projects. Evaluation code and dataset scripts are explicitly marked as not yet released, limiting reproducibility until those are provided.
This repository is a curated literature review (awesome list) focused on the progression from video generation to world models in AI research. It systematically organizes and categorizes recent academic papers (primarily from 2022-2026) into a hierarchical structure covering technical foundations, application domains, evaluation methodologies, and theoretical perspectives. The main value lies in its comprehensive paper collection with structured metadata including publication dates, venues, paper links, project pages, and official code repositories. The repository follows a standard awesome-list format using markdown tables with shields.io badges for easy navigation to paper resources. Content is organized into eight main sections: Generation 1-4 (covering faithfulness, interactiveness, planning, and randomness aspects), Evaluation and Datasets, Study and Rethinking, Downstream Tasks, and World Models for Other Applications. Each section contains subsections detailing specific technical approaches (e.g., GAN-based, U-Net-based, DiT-based video generation) and application areas (robotics, autonomous driving, gaming). The list is actively maintained with frequent updates incorporating very recent pre-print papers. As a static reference resource, the repository contains no executable code, data processing pipelines, or computational components. Its reproducibility is high because the curated content is directly available in the README upon cloning, requiring no setup or dependencies to access the paper list. The organizational pattern demonstrates good practices for academic literature curation, though it does not contain novel reusable data engineering components beyond the standard markdown documentation format.
PopSift is a CUDA-accelerated SIFT (Scale-Invariant Feature Transform) feature extraction library. It provides a C++ library (libpopsift) with an asynchronous API for enqueuing images and retrieving features, and two command-line demo applications (popsift-demo, popsift-match). The library implements the full SIFT pipeline (Gaussian pyramid, extrema detection, orientation assignment, descriptor computation) on NVIDIA GPUs, supporting multiple algorithmic variants (PopSift, VLFeat, OpenCV) for research comparison. It is designed for high-performance real-time feature extraction and can be integrated via CMake. Despite its research origins (Simula Research Laboratory), the codebase is well-structured with modular CUDA kernels, debug utilities, and parallel primitives. The library is actively used in computer vision projects and maintains compatibility with recent CUDA versions while preserving legacy fallbacks.
Diffgram is a full-stack data annotation platform designed for creating, managing, and reviewing annotations on images, video, audio, text, 3D point clouds, and geospatial data. The system is composed of multiple microservices: a Flask-based REST API (`default`), a background processing service (`walrus`) for media ingestion, export generation, and project migration, an event-driven action runner service (`eventhandlers`) that executes automated workflows (e.g., ML model inference, webhooks, exports), and a Vue 2 + Vuetify single-page application frontend. The backend uses PostgreSQL with SQLAlchemy ORM, RabbitMQ for async messaging, and supports multiple cloud storage backends (AWS S3, GCP, Azure, MinIO). Authentication supports OAuth2/Keycloak, API keys, and session cookies with fine-grained role-based permissions. The platform's core domain includes label schemas, attribute groups, instance templates, task/job management with review workflows, discussions/issues, and a custom query engine for filtering files by labels and attributes. It integrates with external AI services (HuggingFace, Vertex AI, Azure Text Analytics) and labeling vendors (Labelbox, Scale AI). The frontend provides rich annotation tools including 2D canvas (boxes, polygons, keypoints), 3D point cloud annotation with Three.js, audio waveform annotation, text token annotation, and geospatial tile annotation. A command pattern enables undo/redo for annotation actions. The codebase demonstrates solid engineering practices with clear separation of concerns, reusable patterns (strategy, observer, command, policy enforcer), and extensive test coverage (Cypress E2E, Jest unit tests, Python unittest). However, there are signs of technical drift: some outdated dependencies, mixed JavaScript/TypeScript in the frontend, and a few legacy modules marked as deprecated. The open-core model is evident from `OPENCORE - ADD` markers, suggesting proprietary extensions. Overall, Diffgram is a mature, actively maintained annotation platform suitable for teams building training data pipelines.
Airtest is a cross-platform UI automation testing framework developed by NetEase, primarily targeting mobile games and apps. It provides a unified Python API and CLI for automating interactions on Android, iOS, Windows, and Linux devices. The core mechanism is image recognition using computer vision algorithms (template matching, keypoint matching with SIFT/KAZE/BRISK, etc.) to locate UI elements, combined with device-specific input simulation via ADB, minitouch, WebDriverAgent, and pywinauto. The framework supports complex gestures, multi-touch, screen recording, and HTML report generation with step logs and screenshots. Architecturally, Airtest is a library with a CLI entry point (`airtest run/report/info`). It uses a device abstraction layer with platform-specific implementations (Android, IOS, Windows, Linux) that share a common API through `airtest.core.api`. The image matching is handled by the `aircv` subpackage, which offers multiple matching strategies configurable via `CVSTRATEGY`. The Android module employs a proxy pattern for screen capture (Minicap, Javacap, ADB) and touch input (Minitouch, Maxtouch, ADB), allowing runtime switching. The framework also integrates with the Poco UI hierarchy inspection tool and supports cloud device testing. While the primary focus is application testing, the repository includes a small benchmark directory that compares the performance of various image matching algorithms, producing metrics like CPU, memory, and keypoint counts. This is a minor component and does not shift the overall classification away from application. The codebase is well-structured with clear separation of concerns, but shows signs of technical debt such as Python 2 compatibility shims, pinned outdated dependencies (opencv-contrib-python ≤ 4.6.0.66, numpy < 2.0), and bundled precompiled binaries (minitouch, minicap, iproxy) that require periodic updates. The project is actively maintained with frequent commits and community engagement, but test coverage is limited due to reliance on real devices.
FAST-Calib is a ROS-based application for LiDAR-camera extrinsic calibration that uses a target-based approach with ArUco markers and circular holes. The system processes synchronized camera images and LiDAR point clouds to detect calibration target features in both modalities, then computes the extrinsic transformation between sensor frames. It supports both solid-state (Livox) and mechanical (Ouster, Hesai) LiDAR sensors through adaptive preprocessing pipelines. The application includes single-scene calibration for immediate results and multi-scene joint calibration for improved accuracy. Key components include modular LiDAR and camera processing pipelines, geometric validation for target detection, and coordinate transformation utilities.
MAGSAC is a C++ library with Python bindings (pymagsac) implementing robust model fitting algorithms for computer vision, including homography, fundamental matrix, essential matrix, rigid transformation, and 2D line estimation. The core algorithm is a template class `MAGSAC<DatumType, ModelEstimator>` that supports sigma-consensus and MAGSAC++ scoring, with interchangeable sampling strategies (uniform, PROSAC, P-NAPSAC, NG-RANSAC, AR-Sampler) and adaptive inlier selection. The Python API mirrors the C++ functionality via pybind11, accepting numpy arrays for point correspondences and returning estimated models with inlier masks. The repository also includes benchmark datasets and Jupyter notebook examples demonstrating usage. While the library is primarily a software package, it contains research-grade data and notebooks, giving it a hybrid character. The code is well-architected with clean separation of concerns, though it has minor maintenance risks such as memory management caveats and limited platform support (Linux/Windows only).
This repository is a collection of C++ sample programs for deploying deep learning models on AXERA AI SoCs (NPU chips). It provides ready-to-run inference examples for a wide range of computer vision tasks including object detection (YOLOv5/6/7/8/9/10/11/26, YOLOX, YOLO-World, PPYOLOE+, RT-DETR, SCRFD), instance segmentation, oriented object detection (OBB), pose estimation, depth estimation, semantic segmentation, stereo matching, and more. Each example is a standalone command-line executable that loads a pre-compiled .axmodel, runs inference via the AX_ENGINE API, post-processes the outputs, and saves a result image. The code is organized per target chip (ax650, ax637, ax620e, ax615) with near-identical copies of examples across platforms, differing only in system initialization and minor chip-specific logic. The architecture is not a library or framework but a set of reference implementations demonstrating the integration of AXERA's NPU SDK. The primary data flow is: load image → preprocess (letterbox resize, color conversion) → push to engine → run synchronous inference → decode model outputs (e.g., anchor-based proposals, DFL, NMS) → draw bounding boxes/keypoints/masks → save output. External dependencies include the AX_ENGINE runtime, OpenCV for image processing, and standard C++ libraries. The build system uses CMake with per-chip toolchain files and GitHub Actions CI for cross-compilation verification. While the code is functional and covers a broad model zoo, it exhibits significant duplication across chip directories and hardcoded configuration (thresholds, image sizes, class names). The monolithic detection.hpp (~3500 lines) bundles many model-specific post-processing functions, increasing maintenance burden. There are no automated tests, and manual memory management is prevalent. Despite these quality issues, the repository serves its purpose as a practical reference for developers integrating AXERA hardware, and it stays current with recent model releases (YOLOv13, YOLO26, Depth Anything V2).
pydegensac is a Python library for robust geometric estimation in computer vision, providing RANSAC-based estimation of fundamental matrices (F) and homographies (H). The core algorithms — DEGENSAC (degeneracy-aware sampling) and PROSAC (progressive sampling) — are implemented in C++ and exposed to Python via pybind11 with Eigen/numpy array interop. The primary deliverable is a pip-installable wheel (pydegensac), built for cp38–cp314 across manylinux2014, manylinux_2_28, macOS, and Windows via GitHub Actions and cibuildwheel. Entry points are the Python API (fundamental matrix and homography estimation) backed by the C++ core exposed through `bindings.cpp`. The codebase is research-grade: the C++ core uses conditional compilation flags (`__HASHING__`, `__D3__`, `__LSQ_BEFORE_LO__`), raw pointers, manual memory management, and commented-out code. Documentation is thin — `bindings.cpp` has placeholder docstrings, and usage examples live in `examples/how-to-use-detailed.ipynb` and benchmark scripts. The repo vendors a full copy of pybind11 (headers, tests, build helpers), which risks drift from upstream. Benchmark infrastructure (`benchmarks/run.py`, `benchmarks/methods.py`) compares against OpenCV RANSAC, MAGAC, Poselib, and Poselib-PROSAC on standard multi-view datasets (IMC/photo-tourism for F, EVD and HPatchesSeq for H), but requires downloading ~1.4 GB of external data from CMP, Hugging Face, and other URLs. Testing relies on golden data (`.npz` files under `tests/data/`) generated by `scripts/make_golden_data.py` from a known-good git HEAD with fixed seeds, plus the bundled pybind11 test suite. There are no unit tests for the C++ core itself. The repo is actively maintained (last commit 2026-09-18, 1 open issue) with wheels published to PyPI, though the publish workflow requires a manual trigger with a build run ID. Known issues include a PROSAC score-orientation bug (benchmark-only, indicating dataset-compatibility risk), platform-specific xfails in the bundled pybind11 tests, and compiler-specific workarounds for GCC/Clang/MSVC. Overall, this is a functional, well-packaged research library with solid CI/CD but notable technical debt: vendored pybind11, placeholder documentation, external data dependencies, and research-code cruft. It is best suited for users needing a fast, reference-quality RANSAC implementation for F/H estimation, or for researchers studying DEGENSAC/PROSAC variants. The hybrid nature is worth noting: while primarily a software library, it includes benchmark/data components (golden data, dataset downloaders, evaluation harnesses) that are integral to its testing story.
face_recognition is a Python library and CLI toolkit that wraps dlib's deep-learning-based face recognition models behind a simple, high-level API. It provides functions to detect face locations (HOG or CNN-based), extract 68-point or 5-point facial landmarks, compute 128-dimensional face encodings, and compare faces via Euclidean distance. The package ships two Click-based CLI entry points — `face_recognition` for batch identification against a known-people folder and `face_detection` for locating faces in images — both supporting multiprocessing and tunable tolerance/upsample parameters. The core logic lives in `face_recognition/api.py`, which is essentially a thin, well-abstracted wrapper around dlib detectors, shape predictors, and the ResNet face encoder, with coordinate conversion helpers to keep the public API in (top, right, bottom, left) tuple form. The repo also includes a large examples directory (webcam recognition, KNN/SVM classifiers, blink detection, batch GPU inference, Flask web service, Raspberry Pi integration) and multi-stage Docker builds for CPU and GPU deployment. Last tagged release is 1.4.0 (2020); the final commit on master dates to mid-2026 but the project shows little active development and many open issues, placing it in 'stale' maintenance status. Technical drift is medium: the codebase still references Python 2.7 in places, pins older dlib versions in some Dockerfiles, and has CI only through 3.10, while the ecosystem has moved to Python 3.12+ and dlib 19.24+. The API wrapper itself is clean and reusable, but the CLI modules and examples are too tightly coupled to the library and filesystem assumptions to be extracted as standalone components.
This repository is a comprehensive PyTorch library for person re-identification (re-ID), providing a full pipeline for training, evaluating, and deploying deep learning models. It includes a modular architecture with data managers (ImageDataManager, VideoDataManager), a large model zoo (OSNet, ResNet, DenseNet, HACNN, etc.), training engines (softmax, triplet, with video support), loss functions (cross-entropy with label smoothing, triplet with hard mining), metrics (CMC, mAP with Cython acceleration), and utilities (feature extraction, model export to ONNX/OpenVINO/TFLite, re-ranking). The codebase is well-organized into subpackages (torchreid.models, torchreid.data, torchreid.engine, etc.) and includes research projects (attribute recognition, deep mutual learning, neural architecture search). While the library is feature-rich and demonstrates strong engineering for its domain, it has not been actively maintained since mid-2022, relies on outdated dependencies (Python 3.6, torch 1.10), and contains deprecated API usage. It remains a valuable reference for re-ID practitioners but requires updates for modern environments.
The code answers the research question of how to customize large text-to-image diffusion models with minimal data and minimal parameter updates, and it compares against DreamBooth and Textual Inversion. It includes reusable components for few-shot personalization, delta checkpoint handling, lightweight model compression, multi-concept weight composition, and CLIP/DINO evaluation, although the codebase shows research-level roughness: duplicated CompVis/diffusers paths, hardcoded paths, and several known bugs in conversion and composition utilities.
This repository is a curated list of research papers and resources on editing techniques for Neural Radiance Fields (NeRF) and 3D Gaussian Splatting. It contains hundreds of entries organized by year and venue, covering seminal works, surveys, and recent advances in 3D scene editing, stylization, inpainting, relighting, and more. The repository is maintained by Weize Li and collaborators from multiple institutions. It serves as a comprehensive index for researchers and practitioners, but contains no executable code, libraries, or data pipelines. It is purely a reference list.
OpenCV Zoo is a curated collection of pre-trained computer vision models optimized for OpenCV's DNN module, accompanied by inference demo scripts (Python and C++), benchmarking harness, evaluation tools, and quantization utilities. The repository serves as a practical resource for deploying models on a wide range of hardware (x86, ARM, RISC-V, NPUs) by providing ready-to-run code for tasks such as face detection/recognition, object detection, segmentation, text detection/recognition, pose estimation, and more. While primarily an application-oriented model zoo, it also includes research-grade evaluation metrics (e.g., LFW, WIDERFace) and quantization workflows, giving it a hybrid character that bridges model deployment and performance characterization. The benchmarking framework and quantization tools are the most architecturally coherent parts; model-specific demos are utilitarian but tightly coupled to their respective ONNX graphs.
This repository is a curated collection of machine learning interview questions and answers, organized by topic (ML fundamentals, deep learning, computer vision, NLP, statistics, coding). It is primarily a documentation site, with a Python build script that generates a static website from the README.md and NLP/README.md files. The site is deployed to GitHub Pages via a CI/CD workflow. The codebase is minimal: the only executable code is the build script, which parses markdown, generates HTML pages with SEO metadata, JSON-LD structured data, sitemap, and robots.txt. The repository is not an application or data pipeline; it is a knowledge base with a static site generator. The build script is well-structured and demonstrates good practices for generating a static site from markdown sources.
This repository is a curated collection of baseline solutions for data science competitions, primarily from Chinese platforms (科大讯飞 AI开发者大赛, Tianchi, DataFountain, AIWIN, etc.) and some international ones (Kaggle). It contains hundreds of Jupyter notebooks and Python scripts covering a wide range of tasks: structured data classification/regression (LightGBM, XGBoost), computer vision (ResNet, EfficientNet, YOLO, CRNN), natural language processing (BERT, TF-IDF, Qwen), time series forecasting, and multimodal challenges. Each baseline is self-contained, providing a quick-start pipeline from data loading to submission generation. The repository serves as a practical learning resource for competition participants, demonstrating common feature engineering techniques, model training loops, and ensemble strategies. However, the code quality is uneven: many notebooks are exploratory with hardcoded paths, outdated library versions, and minimal error handling. The repo lacks centralized dependency management, environment definitions, or modular reusable libraries, making reproducibility low without manual adaptation. Despite these limitations, it offers a valuable catalog of patterns and templates that can be adapted for new competitions.
This repository is a large model zoo (500+ models) focused on converting and quantizing pre-trained computer vision models for edge deployment across multiple frameworks (TFLite, ONNX, OpenVINO, CoreML, TensorRT). It provides hundreds of download scripts for pre-converted model files, conversion pipelines (PyTorch/ TensorFlow to ONNX/TFLite), and a modest set of inference demo scripts for real-time video processing. The repo is primarily a research/engineering resource for edge AI practitioners, not a software application. It exhibits heavy duplication across model directories, relies on an external S3 bucket for model distribution, and uses older framework versions (TensorFlow 1.x/2.x, ONNX opset 11-13). Despite these maintenance concerns, the systematic conversion workflow and post-processing ONNX subgraph generators (e.g., NMS, box decoding) represent reusable patterns for anyone building end-to-end inference pipelines. The repo is hybrid in nature: while classified as data_research due to its focus on pre-trained model artifacts and lack of original training code, it also contains application-like demo code and CLI entry points.