Greg Surma is the personal portfolio website of Greg (Grzegorz) Surma, presenting a number of the author's projects in machine learning, computer vision, autonomous driving, and artificial intelligence. The website itself is a project index; the specific functionality, usage, and limitations of each project must be understood by visiting the corresponding project page, and publicly available materials do not uniformly provide detailed descriptions, pricing models, or platform requirements for each project.
In terms of purpose, this portfolio mainly serves as the author's technical showcase and experiment archive, covering areas such as computer vision, machine learning, iOS development, deep learning, and autonomous driving, with a tech stack that includes Swift, Objective-C, Python, C++, and others.
Its main capabilities can be divided into several categories by project type: image and vision, including Stable Diffusion, AI stock photo collection, artistic style transfer, turning selfies into cartoons, image super-resolution, generative image inpainter, Deep Dream image generator, CNN style transfer, image pixelation, image processing, Holistically-Nested Edge Detection, CNN multi-label classifier, convolutional neural network image classifier, convolutional neural network digit recognizer, CNN tumor recognizer, conditional GAN, DCGAN face generator, DCGAN image generator, making CNNs interpretable, feature visualization, and others; autonomous driving, including autonomous driving AI toy car, autonomous driving depth prediction, autonomous driving lane detection; games and entertainment, including endless AI video loop, fun geography quiz, tic-tac-toe with variations, classic arcade game of math puzzles, minimalist board game, classic 2048 enhanced AI solver, a game combining Tetris and 2048, classic Hex game with an AI opponent, Monte Carlo tree search AI, classic tic-tac-toe with an AI opponent, 3D thinking puzzle board game, and others; iOS/macOS apps, including iPhone/iPad app, iPhone/iPad/mac app, status bar app to prevent Mac from sleeping, macOS tool for custom system screenshots, simple news feed app, GitHub feed app, and others; algorithms and learning environments, including genetic evolution, reinforcement learning, Aaronson Oracle, prisoner's dilemma research environment, Atari 2600 game AI research environment, Snake AI research environment, OpenAI gym environment solver, metaprogramming AI research environment, hyperparameter tuning, transfer learning, character-level text generator, character-level password cracker, Dijkstra shortest path, iterative deepening A*, A* winning path, Sieve of Eratosthenes, public-key protocol, primitive root modulo n, and others; tools and scripts, including a shell script for quickly cleaning Xcode projects, a Python script for displaying system data on an i2c LCD screen, a Python script for downloading metadata and tweets of a specified user, a script for hiding or displaying data in image files, least significant bit algorithm, intelligent hashtag generator, and others.
In terms of usage, the website lists project names in portfolio form. Some projects are labeled with platforms (such as iPhone/iPad, mac) or implementation languages (such as Python Notebook, Swift/ObjC/C++/Python), and some projects are labeled with In-App Purchases, but publicly available materials do not provide unified installation commands, API interfaces, or operating steps.
Applicable scenarios include: browsing the author's collection of AI/ML and computer vision projects, learning about various demonstrative applications, and serving as a reference index for technical practice in related directions.
In terms of limitations: this material is only the content of a portfolio page and does not provide specific functional descriptions, prices, platform compatibility details, or performance promises for each project; the website itself does not constitute a directly usable software product and must be further confirmed based on the specific project pages.
