Alex Naumann

markdownslides. A "markdown-first" approach for generating beautiful browser-based slides easily. Based on Reveal.js and Reveal-Jekyll.

password-protected-website-template. Template to create and host a simple password-protected website with Github Pages.

Plane-Segmentation-Refinement. Segment cuboid-shaped objects (e.g. parcels) without any task specific training data. Code from ICMLA 2020 paper "Refined Plane Segmentation for Cuboid-Shaped Objects by Leveraging Edge Detection".

latex-devcontainer. Devcontainer for using LaTeX in VS Code with auto-formatting and one-click arXiv export and link check.

yaml2pyclass. Code generator that produces a Python class from a YAML input file. Can be used to facilitate code completion for config objects.

synthetic-dataset-generation. Easily create an instance segmentation dataset from an existing pool of objects of interest, distractor objects and background images. Easy configuration, diverse image compositions, multiple blending methods, Dockerized.

CubeRefine-R-CNN. Code for the CubeRefine R-CNN model of our CVPRW '23 paper "Parcel3D: Shape Reconstruction From Single RGB Images for Applications in Transportation Logistics".

conference-deadlines. :alarm_clock: AI conference deadline countdowns + Calendar overview with deadlines and conference dates.

easy-image-scraping. Web application to automatically scrape images from Google, Bing, Baidu and Yahoo.

tampar. Code of our WACV '24 paper "TAMPAR: Visual Tampering Detection for Parcel Logistics in Postal Supply Chains".

pdf-link-checker. Github Action and CLI tool to automatically check all links in a PDF for dead or broken links.

AutoHotKey-GUI-and-Commands. Easy usage of shortcuts for Windows with AutoHotKey. Contains a GUI for quick access to search, folders and more. Complemented with shortcuts for applications and window management.

image-selection-and-cnn-training. Preprocessing of scraped image data (background removal, cropping, ...) and training and evaluation pipeline for a simple CNN using Detectron2.

allocentric_vs_egocentric_pose. This short notebook visualizes and briefly explains the difference between the allocentric and the egocentric pose of an object, as used frequently in machine learning papers.

cv-in-logistics. Literature review website for our paper "Literature Review: Computer Vision Applications in Transportation Logistics and Warehousing".

Logging-Extended. Extension to the standard Python logging package. Facilitates tracking with Git and provides decorators for messages.

a-nau.

python-cuda-envs. Script to automatically map a specific CUDA version to a Conda Python environment.

research-code-template-python.

parcel3d. Project page for the CVPRW 2023 Paper "Parcel3D: Shape Reconstruction from Single RGB Images for Applications in Transportation Logistics".

docker-freecad-cli. FreeCAD compile without GUI

E2EAD.github.io. website of E2EAD

parcel2d. Project page for the ICMLA 2022 Paper "Scrape, Cut, Paste and Learn: Automated Dataset Generation Applied to Parcel Logistics".

codescan. Code of our VISAPP '25 paper "CodeSCAN: ScreenCast ANalysis for Video Programming Tutorials".

COVID-vaccination-availability-check-DE-SL. Selenium based web crawler that checks the availability of COVID19 vaccination appointments for Saarland, Germany

pytorch3d. PyTorch3D is FAIR's library of reusable components for deep learning with 3D data

BRepNet. BRepNet: A topological message passing system for solid models

pytorch-Deep-Learning. Deep Learning (with PyTorch)

ai_chat_mate. 4/13 hackathon on LLM agents @AGI-House: Personal AI Assistant based on Chat GPT for Telegram/Web for hackathon project evaluation.