Volcengine
6 articles found in this topic.
ByteDance's Seedance 1.5 Pro AI Model Generates Dialect Video with Enhanced Audio-Visual Sync
ByteDance's Volcengine unit has launched Seedance 1.5 Pro, an AI video generation model with advanced audio-visual synchronization and strong capabilities in Chinese and dialect content. It supports multi-shot synchronization, generates videos up to 12 seconds, and offers enhanced emotional expressiveness. The model is accessible via Doubao, Jiemeng, and Volcengine Ark.
ByteDance's Volcano Engine Unveils Seedance 1.5 Pro AI Video Generation Model
ByteDance's Volcano Engine has launched Seedance 1.5 Pro, an AI video generation model featuring native audio-visual synchronization, advanced camera movements, and multi-character, multi-language dialogue. This model aims to produce integrated, usable audio-visual works, moving beyond visual-only generation.
ByteDance Unit Unveils Seedance 1.5 Pro With Native Audio-Visual Synchronization
ByteDance's Volcano Engine launched Seedance 1.5 pro, featuring native audio-visual synchronization. It supports 16 dialects and outperforms competitors in instruction adherence and lip-syncing.
Feishu's aily Work Assistant Integrates AI into Core Business Operations
Feishu's aily Work Assistant is transforming work scenarios by integrating AI directly into core business operations. It leverages native platform data and connects with enterprise systems to automate tasks, streamline workflows, and provide proactive assistance, moving beyond general-purpose AI agents.
Tongyi Bailin Releases Upgraded Voice Models, Open-Sourcing Fun-CosyVoice3 and Fun-ASR-Nano
Tongyi Bailin has significantly upgraded its Fun-CosyVoice3 and Fun-ASR voice models, enhancing voice cloning, multilingual support, and speech recognition. They have also open-sourced Fun-CosyVoice3 (0.5B) and Fun-ASR-Nano (0.8B), improving performance in challenging environments and real-time applications.
Volcengine Unveils Doubao-Seedream-4.5 Image Generation Model for Commercial Applications
Volcengine has launched its Doubao-Seedream-4.5 image generation model for public testing, enhancing subject consistency, instruction adherence, and aesthetic quality. It supports diverse commercial applications like advertising, e-commerce, and entertainment, offering advanced multi-image composition and precise instruction execution.