Voxel51

Join us on June 30th for “Beyond Annotation Tools: Building a Complete Physical AI Data Engine with FiftyOne” Register here: https://lnkd.in/gZidbgW3 In this webinar, Nick Lotz will showcase how FiftyOne has evolved into a complete physical AI platform with comprehensive annotation capabilities built directly into the product. In this session, we’ll cover how to: 🏷️ Annotate classification, detection, segmentation, polylines, and video data in one platform 🧠 Build agentic labeling workflows with natural language and visual prompts 🔍 QA annotations in context with embeddings, similarity search, and error analysis 🎥 Run video tracking and instance segmentation workflows for physical AI applications ⚡ Launch scalable auto-labeling pipelines with human-in-the-loop review 🤖 Use AI agents and multimodal workflows directly inside FiftyOne Whether you're building robotics systems, autonomous workflows, video understanding pipelines, or large-scale vision datasets, this session will show how modern teams are evolving beyond standalone annotation tools into complete physical AI data engines. #ComputerVision #AI #MachineLearning #PhysicalAI #Annotation

If you're heading to CVPR this year, come grab a drink with us! Voxel51 and NVIDIA are hosting a happy hour for the Physical AI and CV community, just a 5-min walk from the conference hall. This is a place for fellow researchers and engineers to connect and share ideas. RSVP here: https://lnkd.in/eF7Pc9QG

Join our meetup on July 28 at the InstaDeep offices in #Paris to hear talks from experts on cutting-edge topics across AI, ML, and computer vision. Register here: https://lnkd.in/gi5hViGf Talks will include: • Building Real-World Computer Vision Systems with Voxel51 - Harpreet Sahota 🥑 at Voxel51 • Computer Vision at Nanoscale - Detecting, Segmenting and Analyzing Nanoparticles in microscopic images - Atif Anwer, Ph.D at (Ex) Université Bourgogne Europe • Finetuning VLMs for domain specific tasks - Amine Belhakimi at GoPro #ComputerVision #AI #MachineLearning #VisualAI #PhysicalAI

Join our in-person meetup on June 18 in at Campuslands in #Bucaramanga to hear talks from experts on cutting-edge topics across AI, ML, and computer vision. Register here: https://lnkd.in/g7b9EH2R Talks will include: • Cómo Construir Sistemas de Computer Vision 10x Más Rápido - AdonAI Vera at Voxel51 • Building an AI Sales Assistant in Slack with Google ADK - Carlos Alarcón at Quix • Del Vibe Coding al Desarrollo Guiado por Especficaciones {SDD} - Santiago Carrillo at ADA School #ComputerVision #AI #MachineLearning #VisualAI #PhysicalAI

Join us for a hands-on virtual session on May 27 exploring video-native multimodal AI and how to integrate cutting-edge video understanding models into your computer vision workflows. Register here:https://lnkd.in/ggi3aazq Akshat Shrivastava from Perceptron AI will introduce their latest video-native multimodal model that matches frontier models at a fraction of inference cost, followed by Harpreet Sahota 🥑 Sahota demonstrating how to get started with Perceptron AI inside FiftyOne. #ComputerVision #AI #MachineLearning #VisualAI #PhysicalAI

Join us on August 20th for “From Cold Pool to Hot Queue: Annotation Curation with FiftyOne” — a hands-on workshop with Harpreet Sahota 🥑 on building smarter, more efficient annotation pipelines with data-centric AI workflows. Register here: https://lnkd.in/gMXyHVa2 Most teams spend too much time labeling redundant data while missing the edge cases that actually improve model performance. In this workshop, we’ll cover how to: 🎯 Prioritize the highest-value samples for annotation 🔍 Surface duplicates, outliers, and low-quality data before labeling 🧠 Use embeddings and model signals to curate smarter annotation queues ⚡ Build faster human-in-the-loop workflows with FiftyOne 📈 Improve model quality while reducing annotation overhead Whether you’re building CV systems, managing annotation ops, or scaling multimodal AI pipelines, this session will show practical strategies for turning raw datasets into high-impact training data. #ComputerVision #AI #MachineLearning #MLOps #VisualAI

Join us in Bucaramanga, Colombia on June 19th for the “Data Agents + Visual Agents Hackathon” — a hackathon focused on building the next generation of AI-native workflows with visual AI, agents, and multimodal systems 🇨🇴 Register here: https://lnkd.in/g64JfxuF Hosted by Diego Tarazona Pinzón from Campuslands and AdonAI Vera from Voxel51, we’ll be bringing together developers, AI engineers, builders, and researchers to experiment with: 🤖 AI agents and MCP-powered workflows 👁️ Visual AI and multimodal applications 🧠 Foundation models for image and video understanding ⚡ Rapid prototyping with real-world datasets and tooling 🌎 Collaborative building with the local AI community Construyendo el futuro de la Physical AI juntos. #ComputerVision #AI #MachineLearning #VisualAI #PhysicalAI

Join us on September 2nd for the “Document Visual AI Workshop” — a hands-on session on building modern document AI workflows with FiftyOne. Register here: https://lnkd.in/g4zDn53u Document AI is shifting beyond OCR and text extraction into multimodal workflows that understand layout, visuals, structure, and context together. In this workshop, Harpreet Sahota 🥑 will be covering how to: 📄 Explore and visualize document datasets in FiftyOne 🔍 Identify failure modes and low-quality samples 🧠 Use embeddings and multimodal models for document understanding ⚡ Build evaluation and curation workflows for document AI pipelines 🤖 Accelerate analysis with natural-language agents, MCP Server, and Skills Whether you’re working on forms, PDFs, invoices, enterprise search, or multimodal AI systems, this session will show practical workflows for building more reliable document AI systems. #DocumentAI #ComputerVision #AI #MachineLearning #VisualAI

Join us on June 9th for “Visual AI in Healthcare: Ground Truth in the Foundation Model Era” — a live session on building defensible medical imaging workflows with foundation models and FiftyOne. Register here: https://lnkd.in/gaShfYUE As teams fine-tune models like UNI, MedSAM2, and BiomedCLIP on smaller in-house datasets, label disagreement becomes one of the biggest drivers of model failure — and one of the biggest regulatory risks. In this session, we’ll cover how to: 🩺 Build medical imaging datasets with multiple expert annotations 🔍 Surface reviewer disagreement using embeddings and confidence signals 🧠 Run label-quality screens, near-duplicate detection, and active learning workflows 📈 Fine-tune and evaluate foundation models on auditable datasets 🤖 Use natural-language agents with the FiftyOne MCP Server and Skills Whether you're working in medical imaging, digital pathology, annotation ops, or clinical AI, this session will show practical workflows for improving model quality and trustworthiness. #ComputerVision #AI #MachineLearning #HealthcareAI #VisualAI

Join our virtual meetup on July 23 to hear talks from experts on cutting-edge topics across AI, ML, and computer vision. Register here: https://lnkd.in/gxkgQ3AH Talks will include: • Making Agent Systems Observable, Reliable, and Testable - AdonAI Vera at Voxel51 • Generative AI for Video Trailer Synthesis: From Extractive Heuristics to Autoregressive Creativity - Abhishek Dharmaratnakar at Google • Training-Free Object and Associated Effect Removal in Videos - Saksham Singh Kushwaha at The University of Texas at Dallas • Turning Models into Systems: AI Architecture That Works - Dr. Nikita Golovko at Siemens #ComputerVision #AI #MachineLearning #VisualAI #PhysicalAI