5. Kimi-Audio Kimi-Audio is a new open-source audio foundation model built for universal audio understanding, generation, and speech conversation.
MULTIMODAL AI
-
Comic Translate: AI-powered multilingual manga translation tool
By
–
推荐一个看漫画的神器 – 安装既用 Comic Translate – 软件免费 – 漫画里面的梗懂了 亮点
• 多语言支持：支持从英语、韩语、日语、法语、简体中文、繁体中文、俄语、德语、荷兰语、西班牙语和意大利语之间的翻译。
• 多格式兼容：支持多种漫画文件格式，包括图像（JPEG、PNG）、PDF、EPUB、CBR、CBZ -
Agent S: Open Source Manus Alternative for Task Planning
By
–
推荐 Manus 平替的开源项目 – 任务规划与执行的 Agent Agent S – 免费免费免费 – 更适合 Dev 的 Manus 亮点
• 经验增强的层级规划：结合网页知识和内部记忆，分解任务并动态调整计划
• 多模态大模型驱动：利用图像、文本等多模态信息，实现精准的界面理解与操作
• 广泛的系统兼容性：在 -
Deep Live Cam: Free Real-time Deepfake Face Swapping Library
By
–
推荐一个厉害的库 – live 版 deepfake
— 艾略特 (@elliotchen100) 4 mai 2025
Deep Live Cam – 免费免费免费 – 一键换脸你懂的
亮点
🧠 人脸识别
👀 姿态检测 / 手势识别
🎯 行为分析（举手、倒地检测）
🖥️ 本地运行，隐私友好
🚀 支持 ONNX 加速，轻量高效
场景
🔁 人脸映射
可为多个角色同时使用不同的脸
😂 表情包制作… pic.twitter.com/MCVHDQfDil推荐一个厉害的库 – live 版 deepfake Deep Live Cam – 免费免费免费 – 一键换脸你懂的 亮点 人脸识别 姿态检测 / 手势识别 行为分析（举手、倒地检测） 本地运行，隐私友好 支持 ONNX 加速，轻量高效 场景 人脸映射
可为多个角色同时使用不同的脸 表情包制作 -
Testing GPT-4o and Ideogram v3 for Creative Portrait Sketches
By
–
Trying some 4o prompts with Ideogram v3 > a portrait photo double exposed with a thick white marker single line sketch
-
Gemini and LangChain Integration Guide with Multimodal Capabilities
By
–
Gemini x LangChain Guide Developer guide for integrating Gemini with LangChain, packed with multimodal capabilities, tool calling, and structured output implementations. Includes latest model support and practical code examples. Check out the guide here
-
Prisma: Open Source Mechanistic Interpretability Toolkit for Vision
By
–
Prisma: An Open Source Toolkit for Mechanistic Interpretability in Vision and Prisma is an open-source toolkit that brings powerful mechanistic interpretability tools—previously focused on language models—into the vision and video domain, enabling large-scale analysis and
-
TRAJAN: Motion-Centric Evaluation for Generative Video Models
By
–
Direct Motion Models for Assessing Generated Videos This paper introduces TRAJAN, a new motion-centric evaluation method for generative video models that captures temporal inconsistencies more effectively than metrics like Fréchet Distance (FVD), enabling both
-
X-Fusion Adds Vision to Frozen Language Models
By
–
X-Fusion: Introducing New Modality to Frozen Large Language Models X-Fusion is a dual-tower framework that introduces vision capabilities to pretrained large language models (LLMs) without altering their language weights, enabling unified multimodal understanding and generation.
-
Local AI Generates Animated Product Videos on 6GB VRAM
By
–
Un produit… qui change de couleur, animé en vidéo… par une IA gratuite, qui tourne en local sur 6 Go de VRAM.
— VISION IA (@vision_ia) 3 mai 2025
Pas besoin de Kling. Pas besoin de cloud. Juste vous, votre PC, et des idées.
L’ère du marketing animé génératif a commencé.
Qui va en tirer parti avant les autres… pic.twitter.com/SUzCM8uFZaA product… that changes color, animated in video… by a free AI, running locally on 6 GB of VRAM. No need for Kling. No need for cloud. Just you, your PC, and some ideas. The era of generative animated marketing has begun. Who will make the most of it before the others?