Ritvik Kapila’s Post

𝐄𝐱𝐩𝐥𝐨𝐫𝐢𝐧𝐠 𝐭𝐡𝐞 𝐅𝐮𝐭𝐮𝐫𝐞 𝐨𝐟 𝐀𝐈: 𝐅𝐮𝐥𝐥-𝐁𝐨𝐝𝐲 𝐋𝐚𝐧𝐝𝐦𝐚𝐫𝐤 𝐃𝐞𝐭𝐞𝐜𝐭𝐢𝐨𝐧 𝐰𝐢𝐭𝐡 𝐌𝐞𝐝𝐢𝐚𝐏𝐢𝐩𝐞, 𝐎𝐩𝐞𝐧𝐂𝐕 & 𝐑𝐚𝐬𝐩𝐛𝐞𝐫𝐫𝐲 𝐏𝐢 Currently, I am working on an exciting project that combines MediaPipe and OpenCV with the power of Raspberry Pi to achieve real-time full-body landmark detection. This system can track 33 key body points, enabling precise pose estimation, with potential applications ranging from fitness tracking to rehabilitation assistance. 🔧 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐚𝐥 𝐄𝐧𝐝-𝐭𝐨-𝐄𝐧𝐝 𝐕𝐢𝐬𝐢𝐨𝐧: Our goal is to develop a smart wearable device that leverages this AI technology to provide real-time feedback for improving posture during workouts or assisting patients in physical therapy. The integration of edge computing using Raspberry Pi allows for a portable and cost-effective solution, eliminating the need for cloud connectivity. Although still in the development phase, this project demonstrates how AI-driven systems can transform fitness and healthcare, enabling real-world applications in gyms, rehabilitation centers, and even home environments. I am excited to see how this technology evolves and look forward to creating impactful solutions at PHN Technology Pvt Ltd #AIOnTheEdge #RaspberryPi #FullBodyDetection #FitnessTech #OpenCV #MediaPipe #EdgeComputing #PoseEstimation #PHNTechnology #MachineLearning

Exciting news in wearable technology! Engineers at the University of Maryland have discovered a way to speed up the design of materials for wearables using machine learning and robotics. Traditionally, creating these materials is time-consuming and requires many experiments. But with this new approach, the process is much faster and 95% accurate in predicting material properties. The researchers used titanium nanosheets and natural substances like cellulose (from plants) and gelatin (from animal tissues) to create strong and flexible aerogels.  These materials can be used not only in wearables but also in cleaning up oil spills, storing energy, and insulating windows. Can AI and robotics truly make advanced technology accessible to everyone?  And, how far can we push the boundaries of wearable technology with these new advancements? Let me know what you think! #ai #robotics #advancedtech #technology #artificialintelligence

Scientists from Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR), Bengaluru, an autonomous institute of Department of Science and Technology ventured to create a sensor that not only detects strain but also adapts and “learns” from it. By recreating these pain-like responses, the device paves the way for future smart wearable systems that can help doctors detect stress. Head to our website to read more: https://lnkd.in/gDUnvZhh Jawaharlal Nehru Centre for Advanced Scientific Research (JNCASR) #indiamedtoday #healthcare #healthcarenews #healthcareupdates

*Equal taxpayers Data from wearable sensors (e.g., heart rate, step count) can be used to model mood patterns..... #Exploiting #models #mood #multimodal #Pattern #periodicity #Robustness

Exciting news in wearable technology! Engineers at the University of Maryland have found a way to make designing materials for wearables faster and more efficient by using machine learning and robots. Usually, creating these materials takes a lot of time and many experiments. But with this new method, the process is much quicker and 95% accurate in predicting how the materials will behave. The researchers used materials like titanium nanosheets and natural substances like cellulose (found in plants) and gelatin (from animal tissues) to create strong and flexible aerogels. These advanced materials can be used in wearable technology and other areas like cleaning up oil spills, storing energy, and insulating windows. Can AI and robotics make advanced technology available to everyone?   And, how far can wearable technology go with these new advancements? Share your views! #robotics #technology #smarttech #ai #advancedtech

🌟 Smart Wearable Technology with Single-Cycle EOG Classification 🌟 Excited to share our latest research presented at the 2024 IEEE BECITHCON Conference! 🚀 We’ve achieved 99.26% accuracy in classifying nine distinct eye movements using single-cycle Electrooculogram (EOG) signals. ✨ Highlights of Our Research: ✅ Tackled the challenges of real-time, single-cycle EOG signal processing for precise, non-invasive eye tracking. ✅ Developed a cascaded Artificial Neural Network (ANN) architecture to classify eye gestures, setting new benchmarks in accuracy, precision, recall, and F1-score. 🎯 Why It Matters: This research bridges the gap between theoretical advancements and real-world applications, making low-latency, accurate control of devices via eye movements a reality. The ability to classify complex ocular gestures using AI and with minimal hardware opens doors for revolutionary applications in healthcare and beyond. 💬 Let’s connect and explore collaborations in the areas of biomedical signal processing, wearable technologies, and AI-driven assistive devices. Your thoughts and feedback are most welcome! #BiomedicalEngineering #SmartWearables #EOGClassification #HumanComputerInteraction #AI #DeepLearning #EOG

I’m happy to share our latest review article, "Reshaping the healthcare world by AI-integrated wearable sensors following COVID-19" which has been published in the prestigious Chemical Engineering Journal (CEJ) (Impact Factor: 13.4). In this work, we explore how AI-powered wearable sensors are revolutionizing healthcare, offering innovative solutions for post-COVID-19 challenges. From real-time health monitoring to predictive diagnostics, this article highlights the transformative potential of combining artificial intelligence with wearable technology to improve patient outcomes and reshape healthcare systems. A huge thanks to my co-authors, collaborators who made this possible! 🎉 I look forward to hearing your thoughts and insights! Let’s continue the conversation on how technology can drive meaningful change in healthcare. https://lnkd.in/edWkX53A #AI #WearableSensors #HealthcareInnovation #Research #ChemicalEngineeringJournal #COVID19 #ArtificialIntelligence

Wearable robots have been a revolutionary change for billions of humans in their daily life, enhancing capabilities as grip strength, walking and even stableness. One of the most difficult features is that wearables should infer the user intentions for effectiveness and smoothness. The proposed solution is integrating EMG technology as well as computer vision. The problem is overcoming the challenges related to the safety, and privacy of data collection. What strategies could be implemented to ensure the inclusiveness of data while keeping the wearables effectiveness? Check out the link. https://lnkd.in/emS7t4RD

Embr Labs has been recognized on Fast Company's fourth annual Next Big Things in Tech list. This prestigious list highlights innovative technologies poised to make significant impacts across industries, from education and sustainability to AI and robotics. Embr Labs earned its place on the list for their development of the Embr Wave®, a wrist-worn wearable that uses temperature as a novel method to manage hot flashes, improve sleep, and reduce stress. Their newest upgrade (coming out with their Gen 3) employs AI to detect physiological changes that signal an oncoming hot flash and automatically activate a cooling response. COO and Co-founder Sam Shames expressed pride in the recognition, emphasizing Embr's role in wearable tech innovation. Fast Company's editor-in-chief, Brendan Vaughan, praised the featured technologies for addressing critical global issues and showcasing the potential of innovations to revolutionize industries and improve lives. #HealthTech #WearableTech #MenopauseCare #ProudInvestor #WealthingVCClub

The widespread usage of wearables set the foundations for many new applications that process the wearable sensor data. #Human #Activity #Recognition (#HAR) is a well-studied application that targets to classify the data captured by multiple wearable sensors into one of a predefined set of activity classes. Earlier approaches in the area relied on #Machine #Learning (#ML) classifiers, while many researchers proposed solutions based on #Deep #Neural #Networks (#DNNs), e.g., using #Long #Short-#Term #Memory (#LSTMs) or #Convolution #Neural #Network (#CNNs). Even though the usage of #DNNs for #HAR is a widely explored field, the problem of compressing (thus reducing the computational requirements) the resulting #DNN models has been disregarded by the research community.In this work, we first analyze the impact of the size of the various #CNN components (focusing on the convolution and dense layers) to the model validation performance. As a second step, they focus on two #CNN models with diverse memory characteristics (the lean and the fat model) and they employ three different compression techniques (#Filter-#Based #Pruning –#FBP,– #Low #Rank #Factorization –#LRF,– and #dynamic #range #quantization) assuming various compressibility scenarios. ----Dimitrios Gkountelos, Milad Kokhazadeh, @Charalampos Bournas, @Georgios Keramidas, Vasilios Kelefouras More details can be found at this link: