Research & Education

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."

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.

AstaBench Leaderboard offers a comprehensive platform for viewing and comparing benchmark leaderboards across diverse AI categories. Users can explore performance metrics for models in areas such as literature understanding, code execution, data analysis, and discovery. The tool is hosted on Hugging Face Spaces by AllenAI, providing a centralized location to track and evaluate the advancements in AI model capabilities. It serves as a valuable resource for researchers and developers to assess the effectiveness of different AI systems without requiring any input, simply by browsing the available leaderboards.

Bioclip 2 Demo is an interactive application hosted on Hugging Face Spaces, designed for biological research and data exploration. Users can upload images of plants, animals, or other organisms, and the tool will predict their likely taxonomic rank, such as species, genus, or family. This is achieved using a sophisticated large tree-of-life model. The demo also allows users to supply their own taxonomic tree, offering flexibility for specialized research. It serves as a valuable resource for visualization and understanding biodiversity through image analysis, making advanced biological classification accessible.

AI Phone Leaderboard is a Hugging Face Space that offers a comprehensive leaderboard for evaluating the AI performance of various mobile devices. This tool allows users to analyze benchmark results, providing insights into how different phones stack up in terms of AI capabilities. It is particularly useful for AI enthusiasts, researchers, and mobile developers who need to compare and understand the AI processing power of current mobile technology. The platform is hosted on Hugging Face, leveraging its infrastructure for accessibility and community engagement.

Awesome-LongTailed-Learning is an open-source project offering a comprehensive codebase and a curated list of resources focused on deep long-tailed learning. It features a detailed survey that reviews recent advancements in the field, categorizing existing studies into class re-balancing, information augmentation, and module improvement, further broken down into nine sub-categories. The project also provides empirical analyses of various state-of-the-art methods, evaluating their effectiveness in addressing class imbalance issues. Designed to support the research community, it highlights important applications and promising future research directions, making it an invaluable resource for academics and practitioners alike.

Awesome-Mixture-of-Experts-Papers is a comprehensive, curated reading list dedicated to research in Mixture-of-Experts (MoE) models. This open-source GitHub repository provides an organized collection of papers from recent years, categorized by algorithm, system, and application, and further broken down by publication year. It serves as an invaluable resource for researchers, academics, and students looking to explore the cutting-edge advancements in MoE. The project encourages community contributions, allowing users to add missing papers or fix errors, ensuring the list remains current and accurate. It includes papers from major conferences like ICLR, AAAI, ACL, ICML, and NeurIPS, as well as arXiv preprints, offering a broad overview of the field's evolution.

Labnote provides a comprehensive research note and data management solution tailored for BT (Bio Tech) and NT (Nano Tech) researchers and organizations. The platform aims to enhance research efficiency by allowing researchers to focus on the core aspects of their work. It features Labnote Scholar, an AI research assistant that helps users maximize their research data, and Labnote Preclindoc, a specialized tool for managing all stages of non-clinical research, including automated report generation. Labnote supports both individual researchers and small to enterprise-level research institutions with tools for data management, research collaboration, and AI-driven insights.

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.

awesome-rl is a comprehensive, curated list of resources dedicated to reinforcement learning, designed to support researchers and students in the field. Although no longer actively maintained, it offers a valuable collection of links covering theory, lectures, books, surveys, and foundational papers. The repository also includes applications in game playing, robotics, control, and human-computer interaction, alongside a wide array of codes, tutorials, online demos, and open-source reinforcement learning platforms. This resource serves as an excellent starting point for anyone looking to delve into the complexities of reinforcement learning, providing structured access to key academic materials and practical implementations.

Face_Pytorch offers an open-source implementation of various face recognition algorithms within the PyTorch framework. This project includes well-known algorithms such as ArcFace, CosFace, and SphereFace, providing a comprehensive toolkit for researchers and developers. It supports data preparation for CNN training using datasets like CASIA-WebFace and Cleaned MS-Celeb-1M, aligned by MTCNN. The project also facilitates performance testing on benchmarks like LFW, AgeDB-30, CFP-FP, and MegaFace, with detailed verification results provided for different model types and protocols. It's designed for those looking to implement and evaluate face recognition models, offering flexibility for custom dataset paths and parameters.

Eagle 2.5 is a family of frontier vision-language models (VLMs) developed by NVlabs, specifically engineered for long-context multimodal learning. Unlike many existing VLMs that focus on short-context tasks, Eagle 2.5 excels at challenges like long video comprehension and high-resolution image understanding, providing a generalist framework for both. It supports up to 512 video frames and is trained jointly on image and video data, including the novel Eagle-Video-110K dataset. Key innovations include Information-First Sampling for optimal image and text retention, Progressive Mixed Post-Training for enhanced context length processing, and Diversity-Driven Data Recipe. The model also features significant efficiency and framework optimizations, such as GPU memory optimization and inference acceleration, making it suitable for advanced research and development in multimodal AI.

Transeption IGEM BASISCHINA 2025 is an AI application hosted on Hugging Face Spaces, designed to analyze protein sequences. Users can input a protein sequence and the tool will generate fitness scores for all possible single mutations within that sequence. This data is then presented as a heatmap visualization, providing a clear and intuitive way to understand the impact of various mutations. This tool is particularly useful for researchers and students involved in protein engineering and mutation analysis, offering a streamlined approach to predict and visualize the effects of genetic changes.

Emotion-LLaMA is an advanced open-source AI model designed for multimodal emotion recognition and reasoning, leveraging instruction tuning. It addresses the limitations of traditional single-modality approaches by seamlessly integrating audio, visual, and textual inputs through emotion-specific encoders. The model aligns features into a shared space and employs a modified LLaMA model, significantly enhancing both emotional recognition and reasoning capabilities. It was accepted at NIPS 2024 and has achieved top scores in various challenges, including the MER2024 Challenge. The project also includes the MERR dataset, which contains a large number of coarse-grained and fine-grained annotated samples across diverse emotional categories, enabling models to learn from varied scenarios and generalize to real-world applications.

The Embedded-Engineering-Roadmap is an open-source resource designed to assist both aspiring and current Embedded Systems Engineers in navigating their career path and expanding their skill sets. It offers a comprehensive guide structured into three fundamental areas: Software, Hardware, and Soft Skills. The roadmap provides a curated list of learning resources, categorized by type (books, videos, articles, links) and quality (beginner-friendly, invaluable/comprehensive). It emphasizes hands-on projects as the most effective learning approach and includes links to various project ideas and educational websites. Additionally, it offers guidance on career development, search strategies, and recommended IDEs and VS Code extensions, making it a valuable tool for anyone looking to excel in embedded systems.

EasyNMT is a powerful and user-friendly open-source package designed for state-of-the-art neural machine translation across more than 100 languages. It simplifies the process of machine translation with its easy installation and usage, requiring only a few lines of code to get started. Key features include automatic download of pre-trained models, translation between over 150 languages, automatic language detection for 170+ languages, and support for both sentence and document translation. The tool also offers multi-GPU and multi-process translation capabilities, making it efficient for various workloads. EasyNMT integrates models like Opus-MT, mBART50_m2m, and M2M_100 from Facebook Research, providing a wide range of translation directions and model sizes to suit different needs.

Thai Ai is a mobile application designed to help users learn Thai through personalized AI tutoring. The platform offers interactive audio lessons covering a wide range of topics, from daily life to business communication. Users can practice their Thai skills in realistic conversation scenarios with AI tutors, who can be customized by personality, accent, and region. The app provides powerful learning tools such as real-time corrections, message translation, pronunciation assessment, and bookmarking to enhance the learning experience and help users achieve fluency. It aims to move beyond traditional textbook learning by focusing on practical, conversational skills for everyday situations.

SegmentAnythingin3D (SA3D) is an open-source framework designed for 3D object segmentation within Neural Radiance Fields (NeRFs). It allows users to segment any target object in 3D by providing prompts from a single rendered view. The tool projects 2D segmentation masks onto 3D mask grids via density-guided inverse rendering, iteratively refining the 3D masks. SA3D supports various radiance fields without requiring additional redesign. It offers both point and text prompting options through a GUI, and the entire process for obtaining a target 3D model can be completed rapidly, with recent updates allowing 3D segmentation within seconds using 3D Gaussian Splatting.

Visual Vocabulary is an AI tool designed for learning and exploring data visualization, available as a Hugging Face Space. It offers an intuitive platform to browse a comprehensive visual vocabulary overview through easy-to-use, interactive charts and tables. Users can explore various chart types and their applications without needing any special input, making it accessible for immediate use. The tool aims to enhance understanding of data representation, serving as a valuable resource for anyone interested in data visualization, from students to data scientists. Its interactive nature allows for a hands-on learning experience, making complex data concepts more approachable.

MedMamba is the official code repository for "MedMamba: Vision Mamba for Medical Image Classification." This innovative tool addresses the limitations of traditional CNNs and ViTs in medical image analysis by introducing a novel hybrid basic block called SS-Conv-SSM. This block effectively integrates convolutional layers for local feature extraction with State Space Models (SSMs) to capture long-range dependencies, ensuring efficient modeling of medical images from diverse modalities. MedMamba is designed to provide fewer model parameters and a lower computational burden without sacrificing accuracy, making it suitable for real-world applications with limited computational resources. It has been extensively tested across 16 datasets, ten imaging modalities, and over 400,000 images, demonstrating competitive performance in classifying various medical images.

The WebGL Gaussian Splat Viewer is an interactive application designed for visualizing 3D Gaussian splats directly within a web browser using WebGL technology. Users can easily control the camera through mouse, arrow keys, or touch gestures, enabling seamless navigation and exploration of complex 3D environments. This tool is particularly useful for individuals working with 3D graphics, researchers, and developers who need to inspect and interact with Gaussian splat models. Its web-based nature makes it accessible without requiring specialized software installations, offering a convenient way to share and review 3D content.

WebGPU Depth Anything V2 is an advanced AI tool designed for estimating depth in images. Users can upload an image to generate a detailed depth map, which visually represents the distance of objects within the scene. This tool leverages WebGPU technology, suggesting potential for efficient processing directly within a web browser. It serves as an updated iteration of the original Depth Anything model, likely incorporating improvements in accuracy, performance, or features. This capability is particularly valuable for researchers and developers in computer vision, enabling applications that require precise depth information for tasks such as 3D reconstruction, scene understanding, or robotics.

darknet_ros is a ROS (Robot Operating System) package designed for real-time object detection in camera images, leveraging the You Only Look Once (YOLO) system. It supports YOLO V3 on both GPU and CPU, offering significant speed advantages with CUDA-enabled GPUs. The package comes with pre-trained models capable of detecting objects from VOC and COCO datasets, and also allows users to train and deploy networks with their own custom detection objects. It provides ROS-related parameters for configuring publishers, subscribers, and actions, making it highly adaptable for robotics applications. The tool is open-source and actively maintained by leggedrobotics, providing a robust solution for integrating advanced object detection into robotic systems.

WebGPU Real-time Depth Estimation is an AI tool designed for real-time depth estimation from webcam video, leveraging WebGPU technology. This application provides a dynamic 3D-like view of your surroundings, making it suitable for interactive applications and research in computer vision. Users can adjust parameters such as stream scale and image size to optimize the balance between processing speed and visual detail. This capability is particularly useful for developers and researchers who require rapid depth map generation for their projects, enabling them to explore and implement real-time computer vision solutions efficiently. The tool's focus on real-time performance and adjustable settings makes it a valuable asset for experimental and practical applications in depth sensing.