Research & Education

ResumoGPT is an upcoming AI platform designed for summarizing various file types and web content. While the live website currently displays a "Coming Soon" message, the tool is anticipated to allow users to quickly condense large amounts of text into shorter, more manageable summaries. It aims to be suitable for anyone needing to efficiently extract key information from documents or web pages. Further details regarding its specific features, pricing, and availability are expected to be released once the platform officially launches.

Zotero is a free and intuitive tool designed to streamline the research process for academics and students. It allows users to effortlessly collect research materials from the web with a single click, automatically sensing articles, preprints, news stories, and books. Beyond collection, Zotero provides robust organization features, enabling users to sort items into collections, tag them with keywords, and create dynamic saved searches. A key strength is its citation capability, instantly generating references and bibliographies in over 9,000 styles directly within popular text editors like Word, LibreOffice, and Google Docs. Zotero also offers optional synchronization across devices, ensuring research data, notes, and bibliographic records are always up-to-date and accessible from any web browser. Its collaborative features allow for co-writing papers, distributing course materials, or building shared bibliographies, all at no cost. Developed by an independent, nonprofit organization, Zotero is open source, emphasizing user control over data privacy.

FL-bench is a comprehensive open-source benchmark dedicated to federated learning, providing a robust platform for researchers and developers to evaluate and compare different federated learning algorithms. It supports a wide array of methods, including traditional approaches like FedAvg, FedProx, and SCAFFOLD, as well as personalized FL methods such as pFedSim and FedPer. The benchmark also includes methods for FL domain generalization and differential privacy. Users can easily prepare environments, run experiments, and monitor results using visdom or tensorboard. FL-bench offers extensive customization options for FL methods, datasets, and models, making it a flexible tool for advancing federated learning research. It also supports parallel training via Ray for improved efficiency.

Practical_RL is an open-source course designed to teach reinforcement learning in a practical and accessible way. It is taught on-campus at HSE and YSDA, and maintained to be friendly for online students, offering materials in both English and Russian. The course emphasizes practicality, covering essential techniques, tricks, and heuristics for solving reinforcement learning problems. Each major concept is accompanied by a lab to provide hands-on experience. The curriculum includes topics such as introduction to RL, value-based methods, model-free RL, approximate deep RL, exploration strategies, policy gradient methods, reinforcement learning for sequence models, partially observed MDPs, advanced policy-based methods, and model-based RL. It encourages community contributions through a 'Git-course' manifesto, welcoming pull requests for improvements and new materials.

Voice Crush is an innovative mobile application designed to elevate your voice by crushing unwanted noise and eliminating stuttering from audio recordings. Utilizing state-of-the-art denoising AI, the tool ensures your voice remains clear and prominent, even in challenging acoustic environments. Beyond noise reduction, Voice Crush features an anti-stuttering capability that identifies and edits stutters, filler words, repeats, and awkward pauses, making recorded speech sound more natural and boosting confidence. It's particularly beneficial for individuals on a language-learning journey or anyone needing to send articulate voice messages without the common pitfalls of recording. Made with care in Berlin, Voice Crush aims to improve voice message flow and help users conquer speech impediments.

Tekmatix is an all-in-one business platform designed to help businesses run and scale efficiently by consolidating various tools into a single system. It integrates CRM, marketing automation, course creation, sales funnels, and AI bots. The platform enables users to automate lead generation, manage customer relationships, create and sell online courses and memberships, and streamline email marketing campaigns. Tekmatix also offers features for website and funnel building, social media management, reputation management, and payment processing, aiming to put business operations on auto-pilot with 24/7 support and daily tech training.

TOEIC AI Test Prep Questions, powered by Allen Prep, offers a robust platform for students preparing for the TOEIC exam. The tool provides access to a vast database of thousands of practice questions, each accompanied by a complete rationale to ensure thorough understanding. Designed by perfect-scorers, the content covers all subject areas, helping users avoid surprises on test day. It features performance tracking to identify strengths and weaknesses, and allows for both mock and targeted exams. The platform ensures users always have the latest material with automatic content updates, fostering confidence through consistent practice.

rllab is an open-source framework designed for the development and evaluation of reinforcement learning (RL) algorithms. It offers a comprehensive suite of tools and implementations for a wide range of continuous control tasks, along with several key RL algorithms such as REINFORCE, TRPO, and DDPG. The framework is fully compatible with OpenAI Gym, making it a robust platform for researchers and developers in the RL domain. While rllab itself is no longer under active development, its codebase has been adopted and is actively maintained under the name garage, which offers updated features like TensorFlow support, TensorBoard integration, and new algorithms like PPO.

PokemonRedExperiments is an open-source project dedicated to training reinforcement learning (RL) agents to play the classic game Pokemon Red. It offers a platform for researchers and enthusiasts to experiment with different RL algorithms and observe agent behavior within the game environment. The project includes updated and simplified V2 training scripts, which boast faster training times, reduced memory usage, and improved exploration rewards. A unique feature is the ability to stream training sessions to a shared global game map, allowing for collaborative observation and analysis. Users can also track progress locally via TensorBoard or integrate with Weights & Biases. The project provides detailed setup guides for various operating systems, including Windows, Linux, and MacOS, and supports both interactive play with pretrained models and full model training.

Map Diffusers is an AI tool available as a Hugging Face Space, created by sabman. While its intended functionalities are not currently accessible due to a runtime error, it is categorized as an AI application for content generation. The tool is hosted on the Hugging Face platform, suggesting it leverages machine learning models for its operations. Its current state indicates a technical issue preventing users from interacting with its features. The tool's presence on Hugging Face implies a focus on community-driven ML applications and potential for exploring AI capabilities.

DataSciencePython is a comprehensive GitHub repository designed for individuals interested in data science, natural language processing (NLP), and machine learning. It serves as a central hub for Python-based tutorials, offering a curated collection of resources. The repository organizes content into topic-wise lists, covering various aspects of machine learning and deep learning. Users can find tutorials, code examples, and relevant articles to enhance their understanding and practical skills in these fields.

jsfeat is an open-source JavaScript Computer Vision library designed for developers to explore and implement modern computer vision algorithms using JS/HTML5. The library provides a comprehensive set of features, including custom data structures and essential image processing methods such as grayscale conversion, box blur, Gaussian blur, histogram equalization, Canny edges, and various derivative calculations. It also incorporates a Linear Algebra module for LU, Cholesky, and SVD solvers, along with Eigen Vectors and Values. For advanced applications, jsfeat offers a Multiview module with Affine2D and Homography2D motion kernels, and RANSAC/LMEDS motion estimators. Additionally, it includes feature detectors like Fast Corners, YAPE06, YAPE, and ORB, as well as Lucas-Kanade optical flow and HAAR/BBF object detectors, making it a versatile tool for computer vision development.

Qwen-VL, developed by Alibaba Cloud, is a powerful open-source large vision language model (LVLM) that accepts image, text, and bounding box inputs, and outputs text and bounding boxes. It offers strong performance, significantly surpassing existing open-sourced LVLMs on multiple English evaluation benchmarks. Key features include multi-lingual support for English, Chinese, and multi-lingual conversations, end-to-end recognition of bi-lingual text in images, and multi-image interleaved conversations. It is also the first generalist model to support grounding in Chinese, allowing for bounding box detection through open-domain language expression. The model boasts fine-grained recognition and understanding with a 448x448 resolution, promoting detailed text recognition and document QA.

Lenzy AI offers a comprehensive analysis and visualization of two decades of Hacker News discussions. Utilizing clustering algorithms, the platform identifies and presents key trends, recurring patterns, and community insights from the vast dataset. This tool is designed for researchers and analysts to explore the evolution of technology conversations, pinpoint dominant themes, and understand the collective interests of the developer community over a significant period. It provides an overview of discussed topics, making it valuable for anyone interested in the historical trajectory of tech discourse on Hacker News.

Scenic is an open-source JAX library developed by Google Research, specifically designed for computer vision research with a strong emphasis on attention-based models. It facilitates the development of classification, segmentation, and detection models across multiple modalities, including images, video, audio, and multimodal combinations. The library provides essential boilerplate code for launching experiments, logging, and profiling, alongside optimized training and evaluation loops. Scenic also includes input pipelines for popular vision datasets and a collection of state-of-the-art models and baselines, some developed within Scenic and others reimplemented. Its philosophy prioritizes rapid prototyping and simplicity, encouraging forking and copy-pasting for customization before upstreaming widely useful functionalities.

Lyspeak, despite its name suggesting a language learning tool, functions as an affiliate website in Turkish, focusing on online betting and casino bonuses. The site provides lists of current promotions, such as welcome bonuses, free spins, and no-deposit bonuses, from numerous gambling platforms like Roketbet, Betmoney, Milyar, and Fikstürbet. It aims to guide users to reliable and licensed sites offering these bonuses, detailing different types of bonuses like investment bonuses, loss bonuses, and freebet offers. The platform also includes an FAQ section addressing common questions about bonuses, their usage, and terms and conditions, positioning itself as a resource for individuals interested in online gambling promotions.

Face Mesh Workflow is a tool hosted on Hugging Face Spaces that allows users to upload an image, detect faces within it, and generate a 3D mesh. It offers the flexibility to adjust depth sources and customize the generated mesh using various sliders. The primary output is an OBJ file, which can then be downloaded for further use in other 3D modeling or animation software. This tool is particularly useful for those working with facial recognition, 3D modeling, or anyone needing to create 3D representations of faces from 2D images.

DeepResearch Bench is a comprehensive platform designed for evaluating deep research agents, offering a dynamic leaderboard to track and compare their performance. Users can easily search for specific AI models or filter them by various categories to analyze their scores and effectiveness. A key feature is the ability to conduct side-by-side comparisons of two chosen models, allowing for detailed analysis of their results. This tool is particularly valuable for AI researchers and data scientists who need to assess and understand the capabilities of different deep research agents in a structured and comparative manner, aiding in model selection and performance optimization.

phd-bibliography is an open-source curated collection of academic references focusing on optimal control, reinforcement learning, and motion planning. This GitHub repository provides a structured bibliography covering a wide array of topics including Dynamic Programming, Control Theory, Model Predictive Control, Safe Control, Game Theory, Sequential Learning, Multi-Armed Bandit problems, Black-box Optimization, and various aspects of Reinforcement Learning. It is designed as a valuable resource for researchers, students, and practitioners looking for foundational and advanced literature in these complex domains. The bibliography is organized by topic, making it easy to navigate and find relevant papers, and includes links to specific works like AlphaGo and AlphaZero.

EDGS is a Hugging Face Space by CompVis that offers a simplified approach to 3D Gaussian Splatting. Users can upload a front-facing video or a folder of images of a static scene. The tool then automatically extracts frames, and runs a process to optimize the 3D scene. This tool is designed to improve the efficiency of 3D Gaussian Splatting by eliminating the need for densification, making the process more accessible and streamlined for creating 3D representations from 2D inputs. It provides a practical demonstration of the research outlined in the paper "EDGS: Eliminating Densification for Efficient Convergence of 3DGS."

English Grammar App is a comprehensive platform designed to help users learn and practice English grammar from A1 to C1 CEFR levels. It offers over 100 free lessons and 550+ interactive exercises with instant feedback. A standout feature is the AI Grammar Coach, which provides instant explanations for mistakes, acting as an always-available tutor. Users can stay motivated through gamified elements like daily streaks, 46 badges, and a global leaderboard. The platform also includes daily news stories at various CEFR levels with audio, vocabulary, and exercises, making it a well-rounded tool for English language learners.

dev-conf-replay is an open-source repository that serves as a comprehensive collection of replay links for recent IT seminars and developer conferences in Korea. It organizes video recordings from various events, including those hosted by major IT companies like Naver, Kakao, Line, and Samsung, as well as specialized conferences on AI, Big Data, Cloud, DevOps, Blockchain, Mobile, and Programming Languages. This tool is designed to help developers and IT professionals easily access educational content, stay informed about the latest industry trends, and review past conference sessions at their convenience. The repository is regularly updated with new videos and categorized for easy navigation.

Hyperspectral-Image-Super-Resolution-Benchmark is an open-source collection of resources dedicated to hyperspectral image super-resolution. Curated by Junjun Jiang, this benchmark provides a comprehensive list of techniques and papers for generating high spatial and high spectral resolution images. It covers four main classes of super-resolution: spatiospectral super-resolution (SSSR), spectral super-resolution (SSR), single hyperspectral image super-resolution (SHSR), and multispectral image and hyperspectral image fusion (MHF). The resource includes pioneer work, technique reviews, and recent advancements, often with links to PDF papers and code, making it an invaluable tool for researchers and academics in the field.

DeepEMD offers a PyTorch implementation for few-shot image classification, based on the research paper "DeepEMD: Few-Shot Image Classification with Differentiable Earth Mover's Distance and Structured Classifiers." This tool is designed to address the challenge of learning from limited labeled data by employing the Earth Mover's Distance (EMD) as a metric for structural matching between image regions. It includes a cross-reference mechanism to mitigate issues from cluttered backgrounds and intra-class variations, and supports k-shot classification through a structured fully connected layer. DeepEMD has demonstrated significant performance improvements on benchmarks like miniImageNet, tieredImageNet, FC100, and CUB, without requiring extra training or testing data. The repository provides code for model pre-training, meta-training, and evaluation, along with options for different EMD solvers and model configurations.