VideoAnydoor is a zero-shot video object insertion framework jointly developed by the University of Hong Kong, Alibaba Group's DAMO Academy, Lakeside Lab, and Huazhong University of Science and Technology. It can insert specific objects with high fidelity and precise motion control...
PersonaMagic is an innovative high-fidelity face customization technology that achieves personalized image generation through a phased, conditional text strategy. It learns a series of dynamic embeddings based on a Simple Multilayer Perceptron (MLP) network at specific times...
InstructMove is an instruction-based image editing model jointly developed by the University of Tokyo and Adobe. It learns how to perform image manipulations based on instructions by observing frame-to-frame changes in video. The model is based on a multimodal large-scale language model...
LatentSync is an end-to-end lip-sync framework jointly developed by ByteDance and Beijing Jiaotong University. Based on a latent diffusion model of audio conditions, it requires no intermediate 3D representations or 2D feature points. LatentSync uses Stable Difference...
PsycoLLM is a large-scale Chinese psychological language model developed by the School of Computer Science and Information Engineering at Hefei University of Technology. Trained on high-quality psychological datasets, it enhances the understanding and assessment of mental health issues. The model's data...
Casevo (Cognitive Agents and Social Evolution Simulator) is a joint product launched by the School of Data Science and Intelligent Media Communication and the State Key Laboratory of Media Convergence and Communication at the Communication University of China...
No More Tuning (NMT) is a multi-task learning framework jointly developed by UC Berkeley and Alibaba Group. It addresses the problem of prioritizing different tasks in multi-task learning. NMT transforms the multi-task learning problem into a constraint-based learning problem...
MiniPerplx is an open-source AI search engine that uses the Grok 2.0 model for searching. MiniPerplx offers a free alternative for searching web pages, Twitter posts, research papers, and YouTube videos, among other things...
AddressCLIP is an end-to-end image geolocation model built on CLIP technology, jointly developed by the Institute of Automation, Chinese Academy of Sciences, and Alibaba Cloud. The model can achieve street-level accuracy geolocation from a single photograph, directly predicting the image capture...
TradingAgents is a multi-agent LLM financial trading framework developed by UCLA and MIT, simulating a real-world trading firm environment. TradingAgents integrates multiple agents with different roles and...
CreatiLayout is an innovative layout-to-image (L2I) technology jointly proposed by Fudan University and ByteDance. It utilizes the large-scale layout dataset LayoutSAM, containing 2.7 million images...
RealisHuman is an innovative post-processing framework that refines human body parts, such as hands and faces, in generated images. The framework achieves this through a two-stage approach. First, using the original deformed area as a reference, it generates a more refined version...
Cosmos is NVIDIA's generative world modeling platform, accelerating the development of physical artificial intelligence (AI) systems, particularly in autonomous driving and robotics. Cosmos can accept prompts from text, images, or videos...
UCloud Computing is a GPU computing power rental platform under UCloud, focusing on providing efficient and flexible computing resources for AI applications. It supports on-demand rental to meet the needs of different projects. The platform boasts a rich image community, offering multiple...
KAG (Knowledge Augmented Generation) is a domain-specific knowledge service framework launched by Ant Group. It enhances the question-answering performance of large language models (LLMs) in specific domains based on knowledge augmentation, providing knowledge services for vertical fields...
Gemini Coder is a tool based on Google's Gemini API, Next.js, and Tailwind CSS, used for quickly generating app code. Gemini Coder is based on simple descriptions and combines them with Sandpack for real-time...
AIOpsLab is a prototype framework developed by Microsoft, UC Berkeley, UC Illinois at Urbana-Champaign, Microsoft Research, and other institutions for building, evaluating, and improving AIOps agents for cloud service operations management. AIOpsLab...
VideoRefer, jointly developed by Zhejiang University and Alibaba DAMO Academy, is specifically designed for object perception and reasoning within videos. Based on the spatial-temporal understanding capabilities of Enhanced Video Large Language Models (Video LLMs), the model can...
JoyCaption is an open-source image caption generation tool used to train diffusion models. JoyCaption covers a wide range of image styles, content, ethnicities, genders, and orientations, minimizing filtering to understand various aspects of the world, but...
MajicFlus is an AI model developed by MajicFlus based on the Flux.1 architecture. It focuses on generating high-quality portrait images, excelling at capturing the subtlety and beauty of Asian women. The model integrates multiple technologies to generate lifelike human figures...
CHRONOS is a novel framework for generating news timeline summaries, jointly developed by the Department of Computer Science and Engineering at Shanghai Jiao Tong University, Alibaba Group's Tongyi Lab, and other institutions. It's based on an iterative question-and-answer approach and uses large-scale...
Edicho, developed by the Hong Kong University of Science and Technology, Ant Group, Stanford University, and the Chinese University of Hong Kong, is an image editing method based on a diffusion model that enables consistent editing across multiple images. Edicho uses a training-free approach, requiring no...
Mobile-Agent is a mobile, intelligent agent software entity that can move between nodes on a network to act on behalf of the user or other agents. It can interrupt its current execution and resume on another device, depending on the situation...
VideoRAG is a retrieval-augmented generation technique for long-form video understanding. It helps large video language models (LVLMs) better understand and generate visually aligned auxiliary text by extracting visually aligned auxiliary text from videos...