Voxel51

???? Ready to hit the road with AI? We are! Catch us at https://hubs.ly/Q02ZpJVD0 USA in Dearborn, MI, from Dec. 8-10, where the future of autonomous driving and AI comes to life. Stop by our booth to see how FiftyOne can help you build smarter datasets and drive better model performance. https://hubs.ly/Q02ZpFtt0 #AI #AIInnovation #AutonomousDriving #AutonomousVehicles #VisualAI #TechAD

Most ML engineers focus on model architecture and hyperparameters, while ignoring this critical data issue. ?? Duplicate and near-duplicate images are silently undermining your model’s performance. Watch as Harpreet Sahota ?? breaks down a systematic approach to identify and manage duplicates using modern techniques that go beyond traditional methods. Want to learn more? Check out his "Hands-on Data Centric Visual AI" course here: https://hubs.ly/Q02ZnZsM0 #VisualAI #AI #ML #DataScience #DataCuration #OpenSource

What a whirlwind! ?? Last week, we hit up London, Munich, and Berlin for our 3-city Eurovision tour of meetups and workshops, and wow—you all showed up in full force. We had a blast hacking together on computer vision projects and spreading the word about open-source AI. It’s been inspiring to connect with this amazing community IRL and the ideas and conversations were next-level. A huge thank you to everyone who joined us, asked great questions, and made it all worth it. ?? Missed us this time? Don’t worry—our next meetups are already in the works. Check them out here: https://lnkd.in/gy5ERbVn #ComputerVision #AICommunity #AI #OpenSource #MachineLearning #FiftyOne

Warm welcomes, inspiring conversations, and endless opportunities to grow. That’s what starting at Voxel51 feels like. ?? Paula Ramos, PhD shares her reflections on her first week on the team below. ?? #CompanyCulture #TechCulture #AICommunity #MachineLearning #FiftyOne

Thrilled to support innovators like Fredrik Walterson and the Smart Eye team in shaping the next generation of AI-driven automotive solutions! #Mobility #AutomotiveIndustry #SmartDriving #DriverMonitoring #AI #ML #MachineLearning #FiftyOne

Ever wish you could snap a real-life scene and instantly add it to your dataset? You can—thanks to the magic of FiftyOne and Twilio! ?? Here's what you can do with the Twilio Data Ingestion Plugin: ?? Snap a pic (or a few!) and text them to your Twilio number ?? Open your FiftyOne App, and boom—those images are seamlessly added to your dataset ?? Bonus: Filter by message content to only import the pics you need The best part? This plugin was whipped up in just two hours at a hackathon. Yep, FiftyOne Python plugins make it that simple! Check it out here: https://lnkd.in/e8STt3b6 #Plugins #DataScience #ComputerVision #VisualAI #MachineLearning #Python #FiftyOne

How can pixels power real-world robotics? ?? Brent Griffin explores how inexpensive, pixel-based object representations can help drive mobile manipulation in robotics. ?? Check out the sneak peek below and watch the full presentation here: https://lnkd.in/eGH5snUF #Robotics #Robots #3D #VisualAI #AI #MachineLearning

We work hard and have fun along the way. If you want to join a team of passionate, forward-thinking folks who value flexibility, collaboration, and constant growth, check out our open roles and see how you can be part of Voxel51: https://voxel51.com/jobs/ #CompanyCulture #Collaboration #Engineers #TechCareers #FiftyOne

Tomorrow is the final day of our ECCV Redux series! The series has been packed with insights from top researchers, and Day 4 is set to end on a high note so don't miss these cutting-edge talks. Here's the lineup: ?? Zero-shot Video Anomaly Detection: Leveraging Large Language Models for Rule-Based Reasoning by Yuchen Yang from The Johns Hopkins University ?? Open-Vocabulary 3D Semantic Segmentation with Text-to-Image Diffusion Models by Xiaoyu Zhu from Carnegie Mellon University ?? Join us for the last session here: https://lnkd.in/eui_96ae #ECCV #ComputerVision #AI #MachineLearning #ML #AICommunity

CoTracker3 is here to change the game. This SOTA model doesn’t just handle the tough stuff... It thrives on it. ?? By blending unlabeled real-world videos with a semi-supervised training approach, it bridges the gap between synthetic and real-world data distributions like a pro. Want to get hands-on with CoTracker3? Find everything you need to run inference with the model and parse the output into FiftyOne format from Harpreet Sahota ??: https://hubs.ly/Q02Z2FGV0 #PointTracking #ModelTraining #MachineLearning #ComputerVision #VisualAI #Meta