Welcome to the 【AI Daily】column! Your daily guide to exploring the world of artificial intelligence. We bring you the hottest AI news every day, focusing on developers and helping you understand technology trends and innovative AI applications.
Check out the latest AI products Learn More: https://top.aibase.com/
1. Douyin Integrates with Doubao AI: ByteDance Launches Super AI Ecosystem Strategy
Douyin App is testing the integration of Doubao AI, signifying ByteDance's accelerated push towards an AI ecosystem. By opening access points in the short video interface and message list, Douyin aims to break down barriers between AI products and traffic platforms, enhancing AI capabilities and attracting more user traffic. ByteDance is also advancing multiple AI product lines internally, showcasing a comprehensive AI product ecosystem.
【AiBase Summary:】
📱 Douyin is testing the integration of Doubao AI, opening two strategic access points to enhance AI capabilities and user traffic.
💡 ByteDance is pushing forward multiple AI product lines internally, covering various fields and demonstrating a comprehensive AI ecosystem.
🏆 Internet giants place high importance on the application of AI technology in content ecosystems, ushering in a new era of AI service ecosystems.
2. A Powerful Partnership! JiMeng Integrates with DeepSeek: From Prompts to Paintings in One Step
The collaboration between JiMeng and DeepSeek brings great convenience to AI painting enthusiasts. With DeepSeek's intelligent prompt generation function, users no longer need to worry about creative inspiration. Simply input their needs, and they can obtain precise prompts to generate high-quality images. This innovative collaboration not only improves creative efficiency but also makes the painting process smoother and more enjoyable.
【AiBase Summary:】
🖌️ JiMeng integrates with DeepSeek, providing intelligent prompt generation capabilities to solve user creation challenges.
✨ Users only need to input simple requirements, and DeepSeek can generate detailed and precise prompts, improving creative efficiency.
🌟 DeepSeek generates high-quality prompts that effectively guide AI to generate high-quality images, with positive user feedback.
3. Grok's Voice Mode Opens to All: 11 Modes Launched, Built-in Subtitles Make it an English Learning Tool
xAI's AI assistant, Grok, has officially opened its highly anticipated voice mode to all users, introducing 11 unique voice interaction modes and voice subtitle functionality. This update not only enhances user interaction but also provides a new learning tool for English learners. User response has been enthusiastic. While currently only supporting English, Grok's diverse expression and fluency have received high praise, with anticipation growing for future multilingual support.
【AiBase Summary:】
🎤 Grok's voice mode is now open to all users, offering 11 unique interaction styles, including 2 18+ restricted modes.
📚 The added voice subtitle feature helps users better understand the content, making it a practical tool for English learning.
🌍 Users highly praise Grok's fluency and emotional expression, and look forward to future support for more languages.
4. vivo Restructures, Establishes New AI Department and Shifts Large Model Training to On-Device
vivo recently made significant adjustments to its organizational structure, establishing a new AI department, demonstrating its continued investment and strategic layout in artificial intelligence. The new department will focus on on-device training of large models, and commercialization assessments have been temporarily suspended, reflecting vivo's emphasis on the long-term development of AI technology. Additionally, vivo launched DeepSeek-R1, enhancing the intelligence level of its AI assistant and further improving user experience.
【AiBase Summary:】
🆕 vivo's newly established AI department marks the company's continued investment and strategic layout in artificial intelligence technology.
📉 Due to management intervention, vivo has decided to temporarily suspend commercialization assessments and funding for its AI large model.
🚀 The newly launched DeepSeek-R1 has improved the intelligence level of the AI assistant, significantly improving user experience.
5. New Technology Fast3R: Enables One-Click 3D Reconstruction of Thousands of Images at Astonishing Speed!
Fast3R is an innovative multi-view 3D reconstruction technology that can process up to 1500 images in a single forward pass, significantly improving reconstruction speed. Compared to the traditional DUSt3R method, Fast3R uses a Transformer-based architecture to process view information in parallel, eliminating the complex alignment process, improving inference speed, and reducing error accumulation.
【AiBase Summary:】
🌟 Fast3R technology can process up to 1500 images in a single forward pass, significantly improving 3D reconstruction speed.
⚡ Fast3R's Transformer architecture supports parallel processing, eliminating the complex alignment process of traditional methods.
🚀 Compared to DUSt3R, Fast3R demonstrates significant advantages in time and memory usage, making it suitable for large-scale 3D reconstruction applications.
Details: https://fast3r-3d.github.io/
6. A Nuclear Bomb in Music Creation! DiffRhythm Explodes onto the Scene: 10-Second AI Anthems, Vocals and Accompaniment with One Click!
The arrival of DiffRhythm marks a new era in music creation. It uses diffusion models to achieve end-to-end automatic music generation. Users only need to input lyrics and style to get a complete song in just 10 seconds. DiffRhythm can not only generate accompaniment but also automatically create lyrics that perfectly match the melody, revolutionizing traditional music creation and ushering in a new era of AI music creation.
【AiBase Summary:】
🎤 DiffRhythm uses diffusion models to achieve end-to-end music creation; users only need to input lyrics and style to generate a complete song.
⚡ Generation speed is extremely fast, completing a 4-minute 45-second song in just 10 seconds, 50 times faster than traditional methods.
🎼 Built-in powerful large language model automatically creates lyrics that perfectly match the melody, completely revolutionizing traditional composition methods.
Details: https://huggingface.co/spaces/ASLP-lab/DiffRhythm
7. Microsoft Open-Sources Image Model ART, Capable of Generating Multi-Layer Transparent Images
In the field of image generation, Microsoft researchers' "Anonymous Region Transformer" (ART) technology has revolutionized how users interact with generative models. Through anonymous region layout, ART can directly generate multi-layer transparent images based on global text prompts and introduces a layer-by-layer region cropping mechanism, significantly improving generation efficiency—12 times faster than traditional methods.
【AiBase Summary:】
🌟 ART can directly generate multi-layer transparent images based on global text prompts and anonymous region layout.
⚡️ It uses a layer-by-layer region cropping mechanism, significantly improving image generation efficiency, 12 times faster than traditional methods.
💡 A new high-quality autoencoder supports precise control and generation of multi-layer transparent images, promoting interactive content creation.
Details: https://art-msra.github.io/
8. AI Mind Map Tool MindMapper: Generate Interactive Mind Maps from a Simple Link
In the age of information overload, the Mind Mapper application has become a powerful assistant for organizing thoughts. It can transform user ideas into vivid mind maps. Simply input a URL, YouTube video link, or text prompt to quickly generate interactive mind maps.
【AiBase Summary:】
🖥️ Mind Mapper can transform ideas into vivid mind maps, supporting URL, video link, and text prompt input.
🎨 Using Mermaid.js technology, mind maps are not only beautiful but also have dynamic interactive functions, allowing users to easily access detailed information.
📤 Provides the ability to download mind maps as PNG images, facilitating the sharing of knowledge and inspiration.
Details: https://github.com/misbahsy/MindMapper
9. Lei Jun Appears at the First Representative Channel: Xiaomi Will Apply the Latest AI Technology to All Terminals
At the first "representative channel" of the third session of the 14th National People's Congress, Xiaomi founder Lei Jun answered reporters' questions, delving into Xiaomi's role in manufacturing and its direction in technological innovation. He emphasized that manufacturing is the cornerstone of the nation, and Xiaomi will increase R&D investment, especially in the field of artificial intelligence, to enhance consumers' technological experience and contribute to Chinese-style modernization.
【AiBase Summary:】
🏭 Manufacturing is the foundation of the nation, and Xiaomi will firmly pursue technological innovation and high-end development.
💰 Xiaomi plans to invest 105 billion yuan in R&D from 2021 to 2025, with an estimated 30 billion yuan in 2025, with AI-related businesses accounting for a quarter.
🌍 Xiaomi is committed to applying the latest AI technology to mobile phones, automobiles, and smart homes, enhancing its global market influence.
10. Aispeech Completes A5 Round of Financing, Aiming for New Heights in AI Video Generation
Aispeech recently announced the completion of its A5 round of financing, exclusively invested by Jingya Capital, with financing exceeding 400 million yuan, making it a star enterprise in the AI video generation field. Founded in 2023 by Wang Changhu, former head of visual technology at ByteDance, the company boasts a strong team background and has quickly gained favor with multiple investors. This round of financing will be used to accelerate R&D and attract talent, aiming to build a leading AI video generation large model and its applications.
【AiBase Summary:】
📈 Aispeech completed its A5 round of financing, with financing exceeding 400 million yuan, making it a star enterprise in the AI video generation field.
🌍 Its product PixVerse has over 40 million users, with 15 million monthly active users, demonstrating leading technological capabilities.
💼 The company plans to accelerate model R&D and attract high-end talent, actively expanding B-end enterprise services, facing commercialization challenges.
