Article

Explore our curated collection of AI articles covering artificial intelligence, machine learning, AI tools and more

307 Article•Page 1 / 35

Open source

ShotVerse: Powering Cinematic Camera Control for Text-Driven Multi-Shot Video Creation

ShotVerse: Advancing Cinematic Camera Control with Text-Driven Multi-Shot Video Creation

Open source

MA-EgoQA, a multi-agent first-person video question answering platform

MA-EgoQA, a multi-agent first-person video question answering platform, is built on the EgoLife dataset.

Open source

NVIDIA Nemotron 3 Super, an open-source language model from NVIDIA

NVIDIA Nemotron 3 Super, an open-source language model from NVIDIA, provides powerful agent, inference, and conversational capabilities. It is optimized for collaborative agents and high-volume workloads such as IT ticket automation.

Open source

RL3DEdit: Multi-View Consistent 3D Scene Editing

RL3DEdit: Multi-view Consistent 3D Scene Editing Achieves high-quality 3D editing across various scenarios—including dynamic editing, subject replacement, style transfer, background changes, and challenging scene additions—all in a single forward pass.

Open source

InternVL-U: A Unified Multimodal Model for Understanding, Reasoning, Generation, and Editing

InternVL-U, a unified multimodal model for understanding, reasoning, generation, and editing, with 4B parameters, integrates multimodal understanding, reasoning, image generation, and image editing into a single framework.

Open source

Fish Audio S2 Pro, an Open-Source Text-to-Speech Model

Fish Audio S2 Pro is an open-source text-to-speech model trained on over 10 million hours of audio data across more than 80 languages.

Open source

CARE-Edit: Context-Aware Image Editing

CARE-Edit, Context-aware Image Editing, routes diffusion tokens to four specialized experts (text, mask, reference, base) via a lightweight conditional-aware router.

Open source

WildActor: Unconstrained Identity-Preserving Video Generation

WildActor, a framework for generating human videos conditioned on arbitrary viewpoints, enables unconstrained identity-preserving video generation and unrestricted human video generation under any viewing conditions.