ChartRequest’s Post

Hands-free multi-modal AI glasses that can livestream‼️

Check out the latest news from the CU Medicine Today Fall 2023 edition. Stories include how AI is changing health care, heart surgeons using robots to improve patient outcomes, and more! Read the CU Medicine Today Fall 2023 edition here: https://bit.ly/3B22seV #AI #Robotics #Surgeons

I've finished this short course "Prompt Engineering for Vision Models" led by experts Abby Morgan, Jacques Verré, and Caleb Kaiser at Comet, this course dives deep into the fascinating world of vision models, from image segmentation to object detection and generation. Key Takeaways: 1. Image Segmentation: Mastered using #sam for outlining objects with precision, utilizing both positive and negative prompts to refine the segmentation process. 2. #Object #Detection: Learned how to generate bounding boxes and isolate specific objects within images, enhancing the ability to interact with and manipulate visual data. 3. #Image #Generation: Explored the power of #diffusion #models for generating and editing images. Practiced inpainting techniques by replacing parts of images with generated content, which involves careful tuning of model hyperparameters for optimal results. 4. #Finetuning: Delved into personalizing models using #DreamBooth, and #LoRA enabling the generation of specific visual content tied to custom labels. This course has significantly enriched my understanding of how nuanced prompts can dramatically enhance the capabilities of vision models. Thanks to Andrew Ng, the instructors and everyone involved for this enlightening experience from Comet and DeepLearning.AI community! https://lnkd.in/enbr7K6t #VisionModels #MachineLearning #ImageProcessing #TechEducation #ProfessionalDevelopment #generativeai #diffusionmodels

As part of my PhD research I'm examining the role of memory. Memory is what makes computers work. And us – without memory, the world isn't so much fun. Memory is also how culture functions by remembering what's gone before and building on it. And places have memory too if you have time to notice. Here are some AI generated pictures that explore this correlation. Things get weird fast, I'm warning you. https://lnkd.in/e3jbxHnE

🌟 The Future of Events Industry is Here - Embracing AI! 🚀 As we march forward into the era of technological advancements, one trend is revolutionizing the events industry like never before - Artificial Intelligence (AI)! 📈✨ According to Exploding Topics' latest report on technology trends, AI is reshaping the landscape of events, optimizing processes, and enhancing experiences for both organizers and attendees. From personalized event recommendations to seamless registration and smart event management, AI is bringing a whole new level of efficiency and innovation to the table. 🤖💡 At EventsAir, we are at the forefront of this AI-driven transformation, keeping up with the trends and empowering event planners with cutting-edge tools and technologies. Our platform is designed to streamline every aspect of event planning, from dynamic agenda building to real-time attendee engagement. 🎯🔧 🏆 Join Hundreds of event professionals who trust EventsAir as their go-to event planning platform. Experience the ease of organizing successful, immersive events with the best-in-class technology. Let's shape the future of events together! 🌐🙌 Drop me a message today for a quick confidential chat and how we can help! Use my calendar to book a time that suits you best 📅  #EventsIndustry #ArtificialIntelligence #EventsAir #EventPlanning #InnovationUnleashed

I've just finished the "Getting Started with Mistral" short course at DeepLearning.AI! This insightful course, expertly guided by Sophia Yang, Ph.D., covered essential topics such as #Prompting, #ModelSelection, #FunctionCalling, #RAG from Scratch, and #Chatbot creation. Thanks to the DeepLearning.AI community and Sophia Yang, Ph.D. at Mistral AI for their guidance and ongoing commitment to updating AI short courses. This course has provided valuable insights into effectively using Mistral models, setting the stage for future #AI endeavours. https://lnkd.in/ejXQiq4z #DeepLearningAI #AItraining #ArtificialIntelligence #Mistral #ModelSelection #ChatbotCreation

🩺 Introducing "Apna Vaidya": Your AI-Powered Healthcare Assistant 🚀 Did you know that millions of Indians face challenges in accessing qualified medical advice due to severe shortage of doctors? I come from a remote village in West Bengal, India. The nearest health center with a qualified doctor is 15km away. When I was 12, I had a severe migraine attack but nobody in my family or village ever heard of Migraine. I suffered intolerable pain for a week on medications of regular fever, indigestion and painkillers. On the 8th day, when I was taken to a specialist doctor, a simple medicine quickly relieved me. Even today, this is a huge problem faced by millions of people everyday. While it is bad in big cities, it is almost tragic in villages and remote areas.   The fact that India has got cell phone reception and smartphone penetration even in the most remote villages but doesn't have access to quality healthcare professionals is unacceptable. People generally rely on quack doctors & shopkeepers for medical advice because they can't afford or have access to qualified physicians. But, what if you could have a qualified doctor to answer your medical questions 24/7?   In recent times there has been a lot of chatter about possibilities and wonders of AI. Especially, my friend and mentor, MIT prof, Ashwin Gopinath’s commentary on AI got me thinking - what if we can use the power of AI to solve the huge unmet needs of medical awareness in India. I am not an AI expert; but seeing all the chatter, I decided to give it a go anyway. Surprisingly, it was not as hard as I would have imagined it to be. In my free time and on weekends, with help from Sambarta Ray Barman, Avigyan M. for coding & some guidance from Ashwin, we created an AI powered Telegram chatbot that I feel ready to share with the world:   "Apna Vaidya" (https://lnkd.in/gzgc4Ziq) can understand your medical queries in multiple modes:   ✅ Type out your medical question in any Indian languages.   ✅ Send a voice note describing your symptoms, again in any Indian languages.   ✅ Upload a photo of your prescription, wound, or medication.   It's like having a doctor in your pocket, anytime, anywhere. And the best part? We have tuned it for India, with a deep understanding of local health needs and cultural context.   But to make "Apna Vaidya" truly game-changing, we need your help:   🩺 Doctors & healthcare professionals: We'd love your feedback to make "Apna Vaidya" smarter & more accurate. 🧪 Beta testers: Help us catch bugs & identify areas for improvement. 💡 Everyone: Share ideas on what you may want to make "Apna Vaidya" a more powerful tool for public health.   We truly believe - AI can be a force for good, empowering people with knowledge and making healthcare more accessible. With your support, "Apna Vaidya" can transform millions of lives. This is just the beginning of the AI healthcare revolution. Let's make history together! 💪🏽🚀 #AI4Good #HealthcareAccess #InnovationForIndia

I really enjoyed this short course by DeepLearning.AI titled: "Prompt engineering for vision models". This course is about utilizing prompt engineering, image segmentation and object detection methods in stable diffusion to make very personalized and powerful image generations. Developing a pipeline using OWL-ViT, SAM, and Stable Diffusion to generate segmentation masks and apply diffusion inpainting, all using text inputs. It dives deep into fine-tuning the stable diffusion model with the dreambooth method, LoRA weights update and customized data training, all deployed on the powerful Comet API I'm very excited to use this knowledge in my upcoming generative AI projects! #stablediffusion #lora #dreambooth #cometai #deeplearningai #generativeai