“Audiobox Aesthetics is a model for unified automatic quality assessment for speech, music and sound. Try the demo on @huggingface ➡️ https://x.com/AIatMeta/status/1893009390980170001
Paper page – Distill Any Depth: Distillation Creates a Stronger Monocular Depth Estimator
https://huggingface.co/papers/2502.19204
“Wow! Just tried it with various documents, including a handwritten letter from Claes Oldenburg to Ellen H. Johnson. Impressive results! Will definitely add this to the AI toolkit for journalists https://x.com/fdaudens/status/1894745220094349495
I Used ChatGPT as My CAPTCHA Solver—It Got Weird https://www.makeuseof.com/chatgpt-solve-captcha/
“Extracting structured data from unstructured documents is a huge use case for our customers. We’ve just made it a lot simpler with LlamaExtract, now in public beta! LlamaExtract enables you to: ➡️ Define and customize schemas for data extraction, either programmatically or in https://x.com/llama_index/status/1895164615010722233
“This is so incredible. Hume AI just launched Octave 🎤 “The first LLM for text-to-speech — a model that understands words in context, predicting the right emotions, rhythm, cadence and emphasis, making speech sound more human than ever before,” Have been experimenting with https://x.com/rohanpaul_ai/status/1894837459927867582
“”new” SOTA object detector, and it’s NOT YOLOv12 D-FINE is a model released 3 months ago under Apache-2.0 license; I have no idea how it flew under my radar @onuralpszr thanks for adding it to leaderboard leaderboard link: https://x.com/skalskip92/status/1892497124534747193
gibber link | Devpost https://devpost.com/software/gibber-link
“OlmOCR is a new drop by @allen_ai to parse any PDF 📝🤝 I have fed one of my old master’s notes and it did a great job 💗 It is based on Qwen2VL-7B and works out of the box with transformers, has Apache 2.0 license 🔥 https://x.com/mervenoyann/status/1894422823646409090
olmOCR https://simonwillison.net/2025/Feb/26/olmocr/
“Grok 3’s voice mode has no censorship. It’s quite surprising. Grok Voice Chat with ChatGPT”
https://x.com/arrakis_ai/status/1892858641234993381
“”What do you think, you mechanical piece of shit?” (bis) (bis) (bis) It’s fun but @xai guys you need to work a little bit more at the “keep conversation going” prompt diversity lol” / X https://x.com/giffmana/status/1894310343658151961
“🔍 Agentic Multimodal #RAG with #ColPali, #AmazonNova, #AmazonBedrock & #CrewAI! 📚 @_sumand demo shows a practical approach to building a #multimodal knowledge base by: ✅ Setting up a multimodal knowledge base using ColPali (Vision LLM) ✅ Storing embeddings in a https://x.com/crewAIInc/status/1892955027007819872
“Really interesting situation has returned where free-to-access AI is very close to the frontier. You can get o1 for free through Copilot, Advanced Voice free from ChatGPT, a few free tries at the best coding AI via Claude 3.7 and a very solid free Deep Research through Grok.” / X https://x.com/emollick/status/1894840057170657655
“Plus users will continue to have access to Advanced Voice powered by 4o with the existing daily rate limit, which is more than 5x the free limit, as well as access to video and screensharing in Advanced Voice. And Pro users will continue to have unlimited access to Advanced” / X https://x.com/OpenAI/status/1894495908366356607
“Alibaba’s Tongyi Lab dropped Wan2.1, a suite of advanced AI models for video —Beats SOTA models, generates at 2.5x speed —Excels in complex motion, real-world physics, Chinese & English text —Includes editing tools, video-to-audio, and a 1.3B version https://x.com/rowancheung/status/1894691410856607840
“Starting today, we’re rolling out a version of Advanced Voice powered by GPT-4o mini to give all ChatGPT free users a chance to preview it daily across platforms. The natural conversation pace and tone are similar to the GPT-4o version while being more cost effective to serve.” / X https://x.com/OpenAI/status/1894495906952876101
“Had some fun over the weekend experimenting with how AI, text and audio will make totally new education use cases Built with @elevenlabsio and @lovable_dev – check out https://x.com/rtilleard/status/1891463330608361490
“This week so far from Perplexity: 1. Comet Browser Announcement (signups to waitlist and launch to follow shortly) 2. Deep Research API 3. New Voice Mode with faster responses Not done yet. Tomorrow: another update 🙂 🚢” / X https://x.com/AravSrinivas/status/1894820042816307467
ElevenLabs — Meet Scribe the world’s most accurate ASR model | ElevenLabs https://elevenlabs.io/blog/meet-scribe
“HOLY SHITT, Microsoft dropped an open-source Multimodal (supports Audio, Vision and Text) Phi 4 – MIT licensed! 🔥 > Beats Gemini 2.0 Flash, GPT4o, Whisper, SeamlessM4T v2 > Models on Hugging Face hub, integrated with/ Transformers! Phi-4-Multimodal: > Modalities: Integrates https://x.com/reach_vb/status/1894989136353738882
“🚀 Just launched: A toolkit of 20 free, open-source AI tools for journalists. Create, edit & analyze content without coding skills or subscriptions! Features Whisper transcription, website scraping, local image editing & more in one searchable collection. https://x.com/fdaudens/status/1894103089319186718
“Grok 3 Voice Mode, following repeated, interrupting requests to yell louder, lets out an inhuman 30-second scream, insults me, and hangs up https://x.com/goodside/status/1893932239718691167
“SongGen is joining YuE as an open-source text-to-music (Suno-style) model Feed it a 3s voice sample (optional) → describe your song → write the lyrics 🟰 get a song! https://x.com/multimodalart/status/1892533897537192366
“Say hello to Lovify! ❤️ The ultimate @lovable_dev chrome extension – built with @lovable_dev – to supercharge your workflow! Plan PRDs, debug with voice, supercharge prompts, and ship faster. 💻 Get it here: https://x.com/realtalishaw/status/1890488683763232994
“We introduced Helix at Figure, a new AI Vision-Language-Action model that helps humanoids think like a human would The model lets robots understand voice commands and handle items they’ve never seen before In the limit, Figure is an AI business https://x.com/adcock_brett/status/1893708199447978336
“Grok 3 voice in Romantic Mode, prompt: “Hi” https://x.com/Teknium1/status/1893818697338290484
“OpenAI dropped two major updates for ChatGPT: — Improved Deep Research to Plus, Team, Edu, and Enterprise tiers—offering 10 queries/mo vs the $200 Pro tier’s 120 queries — New GPT-4o mini-powered Advanced Voice for free users https://x.com/rowancheung/status/1894691479118868863
“Today, we’re releasing Octave: the first LLM built for text-to-speech. 🎨Design any voice with a prompt 🎬 Give acting instructions to control emotion and delivery (sarcasm, whispering, etc.) 🛠️Produce long-form content on our Creator Studio Unlike traditional TTS that just https://x.com/hume_ai/status/1894833497824481593
NeRF-3DTalker: Neural Radiance Field with 3D Prior Aided Audio Disentanglement for Talking Head Synthesis | NeRF-3DTalker Neural Radiance Field with 3D Prior Aided Audio Disentanglement for Talking Head Synthesis https://nerf-3dtalker.github.io/NeRF-3Dtalker/
“Github 👨🔧: Generate audiobooks from e-books ————- → Generates `.m4b` audiobooks from `.epub` e-books. → Employs Kokoro-82M text-to-speech model for high-quality voice output. → Supports multiple languages including English, Spanish, French, Hindi, Italian, https://x.com/rohanpaul_ai/status/1893071602704863284
“This is cool: The National Library of Sweden forked @xenovacom’s whisper-web project to run their own model in Swedish. https://x.com/fdaudens/status/1894783482041164083
“Microsoft released Phi-4-multimodal (5.6B parameters) and Phi-4-mini (3.8B) to consolidate speech, vision, and text in one efficient model family, boosting low-latency deployment on resource-limited environments. 🔑 Key Features Phi-4-multimodal handles audio, images, and text https://x.com/rohanpaul_ai/status/1894900923430064291
“Happy Valentine’s Day 🫶 Send special voice notes to your loved ones. Built with @lovable_dev 💛 @supabase 💚 and @elevenlabsio 🖤 https://x.com/thorwebdev/status/1890413160529629628
“Video to Audio is now here in #DreamMachine. To generate sound for your video generations, just select the new “Audio” button. Create with a single click or describe with prompts for more customized direction. Audio is available now in beta for free to all users. https://x.com/LumaLabsAI/status/1894063350666957148
“Checkout this project if you work with audio recordings from Videos, Podcasts or Calls/Interviews! 🎧 Offmute is an open source project that uses @GoogleDeepMind Gemini 2.0 to transcribe, analyze, and summarize meetings, podcast or calls! It generates structured reports using a https://x.com/_philschmid/status/1892599725913768161
“📢 New tool alert: FastRTC makes real-time audio apps in Python actually doable! Automatic voice detection, works with any AI model, and free phone calling integration. Perfect for newsrooms needing quick audio solutions. Get started: “pip install fastrtc”” / X https://x.com/fdaudens/status/1894481745551978498
“The Voice Stack is improving rapidly. Systems that interact with users via speaking and listening will drive many new applications. Over the past year, I’ve been working closely with https://x.com/AndrewYNg/status/1895146310296379419
“Extract reliable PDF text at 1/32 GPT-4o cost using a 7B VLM (Fully open-source) Allen Institute for AI introduced an open-source OCR toolkit called olmOCR that extracts plain text from PDFs at over 3000 tokens/s for about 190 USD per million pages, or 1/32 GPT-4o https://x.com/rohanpaul_ai/status/1894757825408295346
“Are LLMs ready to replace OCR solutions? Yes, the OmniAI OCR Benchmark compared OCR providers against LLMs across accuracy, cost, and latency metrics showing Multimodal LLMs are not only better, they are also cheaper with @GoogleDeepMind Gemini 2.0 Flash offering the best https://x.com/_philschmid/status/1893926592977477751
“OCR had a good run, but LLMs are taking over. The OmniAI OCR Benchmark shows multimodal LLMs crush OCR on accuracy, cost, and speed. And the best deal looks like @GoogleDeepMind’s Gemini 2.0 Flash—cheaper, faster, better. OmniAI OCR Benchmark is an open source benchmark that https://x.com/rohanpaul_ai/status/1893998883833041318
“We’re releasing a new Perplexity Voice Mode: the only experience that can reliably incorporate real time voice and real time information across several languages simultaneously! Update the iOS app and start talking! Android will come soon. https://x.com/AravSrinivas/status/1894792092284789173
microsoft/Magma-8B · Hugging Face https://huggingface.co/microsoft/Magma-8B
Microsoft releases new Phi models optimized for multimodal processing, efficiency – SiliconANGLE https://siliconangle.com/2025/02/26/microsoft-releases-new-phi-models-optimized-multimodal-processing-efficiency/
“Today, we’re launching our first technical report! The report details Helix, a generalist Vision-Language-Action (VLA) model https://x.com/adcock_brett/status/1892579315461599658
“Current visual benchmarks for Large Multimodal Models quickly become saturated despite models’ visual reasoning flaws. This paper introduces ZeroBench, a new visual benchmark designed to be impossible for today’s Large Multimodal Models. 📌 ZeroBench exposes current Large https://x.com/rohanpaul_ai/status/1894622170690851261
“TheoremExplainAgent Towards Multimodal Explanations for LLM Theorem Understanding https://x.com/_akhaliq/status/1894947700019470796
“Phi 4 Multimodal Instruct: https://x.com/reach_vb/status/1894991935456124941
“InterFeedback Unveiling Interactive Intelligence of Large Multimodal Models via Human Feedback https://x.com/_akhaliq/status/1893854697619915119
[2502.14195v1] Bridging Text and Vision: A Multi-View Text-Vision Registration Approach for Cross-Modal Place Recognition https://arxiv.org/abs/2502.14195v1
“How are Vision Language Models trained? 🖼️ @Alibaba_Qwen 2.5-VL demonstrates state-of-the-art performance through dynamic resolution processing, absolute time encoding for videos, and a redesigned Vision Transformer in 3B/7B/72B variants for edge and cloud deployment. 👀 https://x.com/_philschmid/status/1892506190656999925
“Computer vision is really cool https://x.com/bilawalsidhu/status/1894219658955252141
“Turns out this is powered by a 7B parameter onboard vision based LLM… ok this is interesting and sort of expected (had this exact discussion in 2023), difference is now the models are more capable” / X https://x.com/abacaj/status/1892611093802910152
