Content & Design

DigiHuman is an open-source project designed for automatic 3D character animation, leveraging pose estimation and landmark generation techniques. It enables the creation of full-body and facial animations on 3D character models based on camera input. Developed with MediaPipe and Unity3D, MediaPipe generates 3D landmarks for the human body and face, while Unity3D renders the final animation after processing these landmarks. The system supports any 3D models with a Humanoid T-Pose rig and can animate multiple blendShapes for detailed facial expressions. Key features include exporting animations as video files, saving animation data for future use, and filtering MediaPipe outputs for smoother results. The project is a B.Sc thesis from Amirkabir University of Technology (AUT).

Opus is a platform designed to support creative and technical professionals in bringing their ideas to fruition. It offers tools and resources tailored for a diverse audience, including gamers, artists, musicians, and storytellers, enabling them to develop and manage their creative projects effectively. Beyond individual creators, Opus also caters to advertisers, brands, and investors who are actively engaged with creative technology, providing a comprehensive ecosystem for project development and management. The platform aims to offer utilitarian luxury for the digital age, suggesting a blend of practical functionality with a premium user experience.

3dvisionlabs GmbH specializes in cutting-edge 3D sensing technology, offering products like the HemiStereo® DK1 and HemiStereo® NX. Their flagship HemiStereo® DK1 Developer Kit 1 boasts an unparalleled 180°x180° (HxV) field of view, making it the world's widest angle depth sensing camera. This innovative technology is designed to facilitate a new generation of applications requiring advanced spatial perception. The company's solutions are engineered for seamless integration into AI and embedded computing systems, providing crucial data for machine understanding of environments. They also offer the C7 Edge Camera Module, further expanding their product line for various industrial and research applications. 3dvisionlabs aims to enhance the spatial awareness capabilities of machines, pushing the boundaries of what's possible in 3D vision.

Wan 2.2 Animate is a revolutionary online AI tool for character animation, offering advanced features completely free. It allows users to transform any character by generating precise facial expressions and dynamic body movements from performer videos. The platform also excels at character replacement, seamlessly swapping characters while preserving authentic movements, lighting, and environmental context. Powered by cutting-edge AI from Alibaba's Tongyi Lab, Wan 2.2 Animate provides real-time processing, multi-format video support, and batch processing capabilities. It's accessible directly in a web browser without any downloads or registration, making it ideal for academic research, creative experimentation, and content creation.

Hexeum offers a comprehensive store for Twitch overlays and stream graphics, catering to content creators looking to enhance their branding and viewer engagement. The platform provides a wide array of professionally designed templates, animations, and overlay packs, including webcam frames, stream labels, full-screen overlays, and alerts. Users can choose from various styles and game-specific designs, ensuring a consistent and appealing look for their streams. Hexeum's products are high-quality, animated, and seamlessly compatible with leading broadcasting software like OBS, Streamlabs, Twitch Studio, and XSplit. They are also designed for easy installation, customization, and management, supporting streaming across multiple platforms including Twitch, YouTube, Facebook, and Kick. The store emphasizes instant downloads and boasts over six years of experience in crafting stream overlays for hundreds of thousands of streamers.

AI3DGen is an advanced, cloud-based AI 3D model generator that transforms 2D images into high-quality, production-ready 3D models. It leverages sophisticated AI technology to reconstruct geometry and depth from your input, generating detailed PBR textures and clean UV maps. The platform supports multi-format export, allowing users to download models in .GLB, .OBJ, or .STL for seamless integration into software like Blender, Unity, and Unreal. AI3DGen offers a free tier with no signup or credit card required, making 3D content creation accessible to a wide audience. For higher fidelity, priority processing, and commercial rights, users can upgrade to Pro or Ultra plans, which provide enhanced detail and advanced features like customizable seeds and poly count.

AI Baby Podcast: Talking Baby is an innovative Android mobile application designed to bring baby photos to life. Users can easily upload any baby picture, which the AI then processes to create an animated, lip-synced baby avatar. The app allows for a high degree of personalization by enabling users to add any audio file of their choice. This audio is then synchronized with the avatar's movements, effectively creating a unique video podcast where the baby avatar appears to speak. It offers a creative and engaging way for parents and family members to transform static baby pictures into dynamic, memorable video content with personalized voices.

Lexikon AI creates personalized "digital twins" based on your real-life text conversations, allowing you to interact with AI versions of people from your past. Users can upload screen recordings of conversations from any messaging app (Slack, Discord, WhatsApp, Tinder) to train the AI. The tool extracts communication patterns, personality quirks, and inside jokes to create a scarily accurate digital twin. This enables users to get closure on past conversations, rehearse difficult discussions, or explore hypothetical scenarios without real-world consequences. It offers a free plan with limited recording time and a Pro plan for extended chat history and instant responses. Lexikon AI is available on web, Google Play, and the Apple App Store, prioritizing user privacy.

Fifa Tryon Demo is an AI-powered application designed for virtual clothing try-on, enabling users to visualize how different garments might look on a virtual avatar. This tool aims to provide an interactive experience for fashion enthusiasts or designers to experiment with clothing styles without physical try-ons. While the concept is innovative, the current live demo on Hugging Face is experiencing a runtime error, preventing users from interacting with the application. The tool is hosted on Hugging Face Spaces, indicating its potential as a community-driven or open-source project, though its full functionality is presently unavailable.

Sigmawave AI offers a world model engine for physical intelligence, creating photorealistic 3D worlds for training, validating, and deploying AI. It provides end-to-end world model infrastructure, including photorealistic world synthesis, hyperscale data production, continuous validation, and intelligence scenario engineering. The platform generates compliant synthetic datasets for both public sector applications, such as public safety and smart city operations, and enterprise AI development, including manufacturing QA and warehouse robotics. Sigmawave AI helps overcome real-world data limitations by providing privacy-compliant, programmable environments with automated precision labeling and edge case coverage at scale, significantly accelerating AI training and validation.

Latent Video is an AI-powered tool designed for generating videos. It leverages advanced latent space techniques to synthesize video content, offering users a platform to explore and experiment with artificial intelligence in video creation. The tool is hosted on Hugging Face Spaces, making it accessible for those interested in AI video synthesis. It provides a free environment for users to engage with and understand the capabilities of AI in generating visual media.

taichi-nerfs offers a robust PyTorch and Taichi implementation for instant-ngp NeRF training pipelines, designed for researchers and developers working with neural radiance fields. This open-source tool allows users to train NeRF models from preprocessed datasets like Synthetic NeRF and 360_v2, or directly from their own videos. A key differentiator is its ability to deploy NeRF rendering pipelines on mobile devices using Taichi AOT, achieving real-time interactive performance on iOS devices such as iPad Pro and iPhone 14 Pro Max. It also serves as a backend for text-to-3D projects like stable-dreamfusion, expanding its utility in 3D content creation. The project emphasizes performance, with optimizations like half2 optimization for compatible GPUs.

MagiScan is an AI-powered 3D scanning application that allows users to transform real-world objects into high-quality 3D models directly from their smartphone. With over 4 million installs and top rankings on both the App Store and Google Play, it provides a robust ecosystem for 3D creation and augmented reality. Beyond its core scanning capabilities, MagiScan supports AI generation of 3D assets from text or images and offers specialized vertical solutions like AVA for medical-grade scanning, FreightSizer for cargo measurement, Orthio for orthotics design, and ARize for AR commerce. The platform emphasizes its proprietary 3D and AI engine, enabling a seamless spatial workflow for scanning, generation, and AR integration.

spark is an advanced 3D Gaussian Splatting renderer specifically designed for integration with THREE.js, allowing developers to create sophisticated 3D renderings within web applications. It boasts features like integration with the THREE.js rendering pipeline to fuse splat and mesh-based objects, ensuring portability across most devices with WebGL2 support. The renderer is optimized for speed, performing well even on low-powered mobile devices. It supports rendering multiple splat objects with correct sorting and handles various major splat file formats including .PLY, .SPZ, .SPLAT, .KSPLAT, and .SOG. Developers can also render multiple viewpoints simultaneously, transform and edit splats for animation, and perform real-time color editing, displacement, and skeletal animation. A shader graph system allows for dynamic creation and editing of splats on the GPU.

Gradio Blender Bpy is a tool designed to streamline the automation of Blender tasks through the Gradio interface. It enables developers to create interactive 3D applications by simplifying the integration between Blender and Gradio. This allows for the development of web-based interfaces that can control and interact with Blender's powerful 3D capabilities. The tool is hosted on Hugging Face Spaces, providing an accessible platform for deployment and experimentation. While the live website currently indicates a runtime error, its core purpose is to bridge the gap between Blender's robust 3D environment and Gradio's user-friendly web UI creation.

FlipaClip is a versatile 2D animation application designed for both professionals and beginners, enabling users to create engaging cartoons and animations directly on their mobile devices. The app provides a comprehensive suite of drawing tools, an intuitive workflow for adding and managing frames, and support for up to 10 layers per frame, enhancing control and detail. Users can integrate sound effects or custom audio, and even utilize rotoscoping by adding videos or images as animation references. With features like onion skinning, a frames viewer for easy rearrangement, and a grid tool for precise alignment, FlipaClip simplifies the animation process. It also supports sharing creations across popular social media platforms like YouTube, Instagram, and TikTok.

The Thin Plate Spline Motion Model is an AI-powered tool designed for generating animations and motion-based content. Hosted on Hugging Face Spaces, it aims to provide a platform for users to experiment with different animation styles and techniques. While the concept is to enable creative animation generation, the current live instance is encountering a 'Memory limit exceeded' runtime error, preventing its functionality. This indicates that while the tool exists, its current public deployment is not operational, limiting immediate use for animation creation.

Angel is a unique streaming platform that curates and distributes inspiring stories, powered by a community of over 1 million Angel Guild members. Members can stream a variety of shows, get free tickets to theatrical releases, and receive early access to content and merchandise discounts. The platform offers different membership tiers, including options for 1080p and 4K streaming, ad-free viewing, and the ability to vote on concepts and films for future greenlighting. Angel aims to support filmmakers and deliver stories focused on courage, redemption, hope, and family strength, with new titles planned for release weekly.

FocusOnDepth is an AI tool designed for depth estimation in images, hosted as a Hugging Face Space. While the tool aims to provide capabilities for analyzing and processing images to determine depth, it is currently experiencing runtime errors due to insufficient hardware capacity. This makes it unavailable for immediate use. When operational, it would be suitable for researchers and developers interested in image processing and AI model testing, particularly those working with depth perception in computer vision applications. The tool is free to use, making it accessible for experimentation and academic purposes.

Stable Fast 3D is an innovative AI tool hosted on Hugging Face Spaces by Stability AI, designed to transform 2D images into textured 3D models. Users can simply upload a picture, with or without a background, and the application will generate a viewable and downloadable 3D mesh. The tool offers several customization options, allowing users to tweak parameters such as foreground size, mesh detail, and texture resolution. Additionally, it provides the capability to add HDR lighting for enhanced visual quality. This makes Stable Fast 3D an accessible solution for quickly creating 3D assets from existing images, suitable for various creative and prototyping needs.

Text3d is an AI tool designed to generate 3D models directly from text descriptions. This platform, hosted on Hugging Face, enables users to create diverse 3D content by simply providing natural language prompts. While the current live website indicates a build error, the tool's core functionality is centered around transforming textual input into three-dimensional assets. This capability aims to simplify the 3D content creation process, making it accessible to users who may not have extensive experience with traditional 3D modeling software. The tool is intended to be a free resource, leveraging AI to bridge the gap between conceptual ideas and tangible 3D representations.

2d-gaussian-splatting provides an official implementation for creating geometrically accurate radiance fields using 2D Gaussian Splatting. This open-source project represents scenes with 2D oriented disks and utilizes perspective-correct differentiable rasterization. It includes regularizations to enhance reconstruction quality and offers various meshing approaches for Gaussian splatting, including both bounded and unbounded mesh extraction. The tool supports COLMAP and NeRF Synthetic datasets, and provides scripts for training, rendering, and evaluation of novel view synthesis and geometric reconstruction. It also features integrations with community resources like WebGL/Three.js viewers and offers performance improvements through CUDA operator fusing.

DarkPose is an open-source project that introduces a novel Distribution-Aware Coordinate Representation of Keypoint (DARK) method for human pose estimation. This method acts as a model-agnostic plug-in, designed to significantly boost the performance of various existing state-of-the-art human pose estimation models. It has demonstrated impressive results, including achieving 76.4 on the COCO test-challenge (2nd place entry of COCO Keypoints Challenge ICCV 2019) and being accepted by CVPR2020. The project provides detailed results on COCO val2017, COCO test-dev2017, and MPII val datasets, showcasing its effectiveness across different benchmarks. DarkPose is particularly valuable for researchers and developers working on computer vision tasks requiring precise human pose analysis.

dn-splatter is an open-source project that integrates research papers (DN-Splatter and AGS-Mesh) to enhance Gaussian splatting models with depth and normal supervision. This leads to improved novel-view synthesis and more accurate mesh reconstruction, particularly when using data captured from smartphones (like iPhones). The tool provides pipelines for both DN-Splatter and AGS-Mesh, with AGS-Mesh offering advancements in depth and normal filtering strategies for better mesh quality. It supports various depth loss types, monocular or rendered depth supervision, and includes scripts for data preparation, such as generating pseudo ground truth normal maps and aligning monocular depth estimates. The project is compatible with Nerfstudio environments and offers installation via Conda/Pip or Pixi.