This week’s look at AI and ML news includes: A Fresh Angle For Healthy Eating Envisioned by Computer Vision and ChatGPT, Research To Reconstruct 3D Images With Eyeball Reflections, AI May Change The Way We Fall in Love, and Plainsight Announces U.S. Patent for Foundational Vision AI Technology
Diet Super Vision with a Helping of ChatGPT-Inspired Meals
SnapCalorie Sees A Fresh Angle for Easing Tedious Calorie- and Nutrition-Tracking
SnapCalorie tracks the calorie count and macronutrient breakdown of meals from a single photo taken with a smartphone. The company has emerged in a crowded category of diet apps and calorie-counting services as a funded player with a recent $2M raise from investors including Accel, Index Ventures, former CrossFit CEO Eric Roza, and Y Combinator.
Depth sensing obtains the 3D shape and appearance of objects in imagery.
In addition to food detection and classification with caloric estimations, SnapCalorie is taking a fresh approach using computer vision-enabled depth sensors on supported devices to measure portion sizes. Portion sizing is where inaccuracies universally occur–both by humans and technology. With depth sensing, SnapCalorie is able to obtain the 3D shape and appearance of food items in images. Depth sensing is also used in AR/VR applications allowing virtual objects to be placed in the correct locations in AR/VR environments. While their algorithms outperform human estimation, the company states they also employ a team of human reviewers to improve accuracy.
The SnapCalorie team led by Wade Norris, former Google Lens co-founder, are solving very complex computer vision problems to better inform health-conscious consumers. As the company works to expand their dataset of 5,000 meals, including thousands of photos of each meal and weights of every ingredient and a current bias toward American cuisine, they’ve integrated a chatbot powered by ChatGPT into their app for meal suggestions. Bon appétit!
Reflections of The Way Life Used to Be
Researchers use eyeball reflections to recreate 3D images of scenes
Researchers from the University of Maryland, College Park have developed a computer vision model that can reconstruct 3D images of a scene from the reflections on a person’s eyeballs. The system, which is based on a technique called neural radiance fields (NeRF), takes between five and 15 digital photographs from different angles of an individual’s face while they look at a scene, and then extrapolates from the reflections in their eyes to create a 3D model of the scene.
The method is still under development, but it has already produced “reasonable” results in replicating real-life objects. However, the images are blurry because of the difficulty of rendering the shape of the cornea, the clear outer layer at the front of the eye. Researchers believe a number of applications can result, such as security, augmented reality, and virtual reality. For example, it could be used to create 3D models of people’s faces for identification purposes, or to generate virtual worlds that are more realistic and immersive.
Majority of Dating App Users Swipe Right for AI
Increased accuracy in connecting people based on personality traits and interests intended to revolutionize online matchmaking
Online dating has become part of our culture, as people of all ages have turned to the internet to find love and companionship. While many people succeed, there are many who are frustrated with the current state of online dating. They feel that it’s too superficial and unable to help them to find truly compatible partners.
A recent study by the Pew Research Center found that 57% of online daters believe that AI could be used to improve the dating experience. So, Lior Baruch, the co-founder and CEO of AlgoAI Tech, is not alone in believing that AI can be used to revolutionize online dating. Baruch says that his company has developed an AI-powered app that can match users based on their personality traits and interests. While still in development, the app is more accurate than traditional dating apps, which often rely on superficial factors such as physical appearance.
AlgoAI’s algorithm adjusts its matching methods based on what is or isn’t working for users and the app’s AI robot named “Lora,” adapts a user’s experience based on their personality and preferences, as well as deep psychological and behavioral analysis.
Of course, there are challenges that need to be overcome before AI can be widely adopted in online dating. For example, these types of AI algorithms need to be trained on large datasets of user data, and this data can be difficult and expensive to collect. Ongoing responsible model oversight will also be needed.
We’re rooting for online daters to find love as AI efficiency and effectiveness conquers all and proves that it’s here for the right reasons.
Plainsight Granted U.S. Patent for Foundational Vision AI Technology
This week at Plainsight, we announced our U.S. patent for foundational vision AI technology for identifying, counting, and monitoring livestock in real time.
For some of the world’s largest food producers and processors, we’re leveraging our proprietary technology for custom development of applications tailored to each customers’ needs and environment. Our proprietary technology, expertise, and end-to-end vision AI platform helps us solve enterprise challenges faster with less data, achieving near-perfect accuracy at lower cost. Check out our full press release to learn more and request a call with our experts to discuss your challenges and the ways we can help.
About Plainsight
Plainsight provides the unique combination of AI strategy, a vision AI platform, and deep learning expertise to develop, implement, and oversee transformative computer vision solutions for enterprises. Through the widest breadth of managed services and a vision AI platform for centralized processes and standardized pipelines, Plainsight makes computer vision repeatable and accountable across all enterprise vision AI initiatives. Plainsight solves problems where others have failed and empowers businesses across industries to realize the full potential of their visual data with the lowest barriers to production, fastest value generation, and monitoring for long-term success. For more information, visit plainsight.ai.
