AREA 23 | An IPG Health Company’s Post

I’m excited to share that I presented our paper, "Robot Design for Social Intervenability" at the NordiCHI 2024 conference last week during the "AI and Robots" session. The work is co-authored by Jörg Pohle (Alexander von Humboldt Institute for Internet and Society). Curious why we believe designers should put strong emphasis in creating robot systems that allow humans to socially intervene in the robot’s behaviors, rather than focusing on optimizing context understanding for flawless operation? Have a look here: https://lnkd.in/ee5cmVc7 #NordiCHI2024 #nordichi #RobotInteraction #SocialRobots #CooperativeIntelligence

Human I/O is a unified approach that uses egocentric vision, multimodal sensing, and LLM reasoning to detect situational impairments and assess a user’s ability to interact with their hands, vision, hearing or speech in a given situation. #Google #GenAI #LLM #googleresearch #humanIO https://lnkd.in/gttBJvwK

Enhancing Visual Realism: NeRF-Casting’s Leap in Specular Reflections 💡 We're stepping into a future where the clarity and consistency of reflections in virtual scenes are nothing short of photorealistic. NeRF-Casting offers an innovative leap forward, addressing longstanding challenges in rendering realistic view-dependent appearances, especially for specular objects. → High-Quality Reflections: NeRF-Casting outshines previous methods with superior reflection rendering of both distant and nearby content. → Efficient Processing: By integrating ray tracing, it avoids the computational heft of traditional Neural Radiance Fields, enabling faster optimization without sacrificing detail. → Advanced Modeling: Combines a minimalistic MLP for color decoding with advanced sampling techniques to deliver high fidelity visuals. → Real-World Application: Unique among NeRF techniques, it is specifically effective in real-world scenarios, balancing performance with spectacular visual output. Discover more about how NeRF-Casting is setting new standards in digital imaging technology here - https://lnkd.in/g__YaunU #DigitalImaging #NeRF #VisualRealism -- Post curated by Vishal Pandey. Reach out to him for future features.

I am happy 🎉 to share that our work "Tell What You Hear From What You See - Video to Audio Generation Through Text" by me, Kun Su, and my advisor Eli Shlizerman as been accepted by NeurIPS 2024 as poster. In this paper, we propose a multi-modal generative framework that takes video and an optional text prompt as input, and generates audio and optional textual description (caption) of the audio. At its core, the model consists of an instruction fine-tuned LLM with a projection layer that maps video features to the LLM vector space, and a bi-directional transformer that generates audio tokens from visual frames and optional text prompt using iterative parallel decoding.  Such a framework has two unique advantages: i) Video-to-Audio generation process can be refined and controlled via text that complements the video context, and ii) The model can suggest what audio to generate for the video by generating audio captions. This unlocks novel applications such as text-guided video-to-audio generation and video-to-audio captioning.  To check generated samples, please visit the https://lnkd.in/gJEdHJj3 (best view with Chrome browser). The camera-ready full paper will be coming soon, see you folks again in Vancouver! #NeurIPS #NeurIPS2024 #multimodalml #llm #genai

Turn yourself to Mr.Bean by GenAI tools in Lens Studio 5.0 https://lnkd.in/eku-4-9y

🤖 AI as a Companion? 💬 Imagine wearing a necklace that not only listens but also talks back, offering companionship and emotional support throughout your day. That’s exactly what the Friend AI Necklace promises. It’s designed to help reduce feelings of loneliness and support those with social anxiety by engaging in real-time conversations and sending thoughtful, personalized texts. 🧐 Would you wear this? Do you think AI could really replace or enhance human interaction when it comes to emotional well-being? Is this the future of personal tech, or does it cross a line? Watch the video below and let me know your thoughts: 🔹 Would you give it a try? 🔹 Where do you see this fitting into our daily lives? 🔹 How do you feel about AI’s role in offering emotional support? I’m genuinely curious to hear your perspectives! Drop a comment below. https://lnkd.in/gS46f5va Holli Wagner Greg Wasmuth Amanda Fornal Jeanne Lasheff Jessica H. Brad Parks Evan B. Stone Kamil Banc Gary Schare WITI #AI #WearableTech #CompanionAI #TechInnovation #EmotionalSupport #FutureOfTech

🌟 Excited to share the latest blog post on EchoWrist, a groundbreaking technology for hand pose tracking and interaction recognition! Our hands play a crucial role in our interactions with the world, making understanding hand poses and interactions vital for human-computer interaction. In this post, we introduce EchoWrist, a low-power wristband that leverages active acoustic sensing to continuously estimate 3D hand poses and recognize hand-object interactions. Equipped with speakers emitting inaudible sound waves and sophisticated microphones, EchoWrist captures rich information about hand shape and interaction contexts. Through deep learning inference, it efficiently identifies various hand activities with remarkable accuracy. Results from user studies demonstrate EchoWrist's effectiveness in tracking 3D hand poses and recognizing hand-object interactions, operating at an impressive 57.9mW power consumption. Read the full details at: https://bit.ly/3SiSuk5 #HandInteraction #AcousticSensing #TechnologyInnovation

Title: Hand Gesture Recognition using MediaPipe Studio _Matter:_ In this project, we leverage MediaPipe Studio's pre-built Hand Gesture Recognition project to develop an application that accurately identifies hand gestures. We demonstrate how to: - Configure the project settings for optimal performance - Run the project and test its recognition capabilities - Refine the model by adjusting parameters and retraining for improved accuracy This project showcases the effectiveness of MediaPipe Studio in enabling hand gesture recognition, paving the way for innovative applications in various fields. Let me know if you need any further assistance! AIMER Society - Artificial Intelligence Medical and Engineering Researchers Society #mediapipe #aimers #handgesture

💡 Here's how you can assess how well your model memorizes your data: 🚨 Consider the risks of even a minor error in critical domains like healthcare or finance. Despite rigorous prompt-tuning, retrieval-augmented generation, and fine-tuning, LLMs may still falter, making critical precision errors. 🔍 Introducing the Lamini LLM Photographic Memory Evaluation Suite! This innovative suite rigorously tests LLMs on tasks requiring photographic memory for dependable and precise model evaluation. 🧠 Why is this crucial? Errors as small as a single letter in a medical code or a slight miscalculation in financial data can have dire consequences. 📊 With benchmarks ranging from MMLU to TruthfulQA, their suite pushes LLMs to their limits in sectors like e-commerce, medicine, and finance. 🏆 Track model performance on their leaderboard: https://lnkd.in/d6EsnbWD 📈 Discover how they challenge LLMs to maintain absolute accuracy in critical scenarios: https://lnkd.in/dQ9u2jgY #AI #MachineLearning #HealthcareAI #FinanceAI #TechInnovation #DataAccuracy

Past Visions Reimagined: Xonception MMXXIV I have revisited some of my decades-old digital art with Generative AI to breathe new life into it. One example of such a reimagination is this visit to a synthesised landscape of alien technology and biomechanical creatures. #AIart #digitalart #GenerativeAI #PastVisionsReimagined Software: Adobe Lightroom, Adobe Photoshop, Adobe Premiere, Midjourney, Strata 3D