AI related news

Regular updates on major news in the AI field for easy browsing!

2024.1.2

🎙️ OpenVoice - A multifunctional instant voice cloning tool:

• Accurately clones reference timbres, supports multiple languages and accents;
• Flexible control over voice style, including emotion, accent, rhythm, etc.;
• Zero-sample cross-language voice cloning capability;

Experience address: OpenVoice

🎞️ FlowVid - Video to video synthesis tool:

• Enhances the temporal consistency between video frames.
• Supports various video editing functions, including style transformation, object replacement, etc.
• Can be combined with image to image editing models.

2024.1.3

🐭 Mickey-1928: A model focused on generating images of Mickey Mouse:

• Fine-tuned based on Stable-Diffusion-xl, generating images of Mickey Mouse in the style of 1928.
• Training data includes still frames from three cartoons.

Model address: Mickey-1928

🗣️ DreamTalk: Open-source software for animating portrait photos according to audio

• Makes portrait photos speak or sing according to audio.
• Keeps mouth shapes and expressions consistent.

GitHub address: DreamTalk

2024.1.5

📺 AI Tube: The first AI video platform

• All videos are completely generated by AI.
• Various types of video channels, such as music, animation, games, etc.

2024.1.6

🤖 ChatGPT Wrapper Open Source Collection:

• A collection of wrapper programs for ChatGPT, Midjourney, SD, and WeChat bots.
• Provides a one-stop guide, covering FAQs and basic strategies.
• Suitable for beginners to set up and operate AI sites.

GitHub address: Wrapper Open Source Collection

2024.1.7

📘 Stanford University develops WikiChat:

• Based on Wikipedia information, high accuracy.
• Almost no hallucinations, highly conversational.
• Adaptable to various query and dialogue scenarios, high performance.

Experience address: WikiChat

💻 Copilot-GPT4-Service: Free use of GPT-4:

• Use GPT-4 through GitHub Copilot requests.
• Free and unlimited use of GPT-4 model.

GitHub address: Copilot-GPT4-Service

🔊 OpenAI upgrades GPTs to support voice conversations:

• GPTs now have voice conversation capabilities.
• The launch animation has also been updated.

2024.1.11

🌐 OpenAI launches GPTs Store and ChatGPT Team subscription plans:

• 3 million GPTs were created.
• Provides a private GPT store section.
• Featured GPTs and team plans.

Address: GPTs

🤖 Chatbot UI: Open-source web UI framework for chatbots:

• Supports integration with various AI models.
• Fully featured, 100% open source.

GitHub address: Chatbot UI

2024.1.12

🚀 Key points from Ultraman’s speech at YC W24:

• GPT-5 might achieve an exponential leap, bringing challenges.
• OpenAI API will become faster, more reliable, and cheaper.
• Advised against focusing on overcoming GPT-4 limitations.

2024.1.14

🔍 Surya: Multilingual document OCR tool:

• Provides accurate line-by-line text detection and recognition.
• Features: Line-by-line text detection, text recognition, table and chart detection (coming soon).
• Supported languages: Includes English, Chinese, Japanese, Hindi, etc.

GitHub address: Surya</