Skip to yearly menu bar Skip to main content


CVPR 2025 News

 CVPR 2025 Exhibitor/Sponsor News

Onsite News

  • 3LC (Booth 1309) - 3LC is a data centric platform that gives you insight at a sample level on how your training data is affecting your AI model. The real differentiator between companies these days is their proprietary data, and not the AI model. Come see us at both #1309 for more details!

  • Bayer (Booth 1310) - Bayer is here but “NOT here!” – check out Bayer’s booth 1310 and learn about the upcoming launch of their AI Innovation Platform, powered by Google Cloud technology. The platform offers healthcare organizations and startups a secure development environment, advanced data processing tools, AI-assisted regulatory document creation and flexible deployment options. Streamline the development of AI-driven medical imaging software from concept to market launch. Find out more about how Bayer are accelerating innovation for medical imaging at their booth 1310.                                                                                                                                                                                                                                                                                                                                          
  • Tiltan (Booth 1619) - Tiltan's advanced, innovative solutions are perfectly suited for the needs of HLS, defense and first responders. When it comes to first responders, Majestic.ai datasets can assist in the development of AI models to identify objects such as fires, people in distress, anomalies, vehicles under flood water as well as individuals on roof tops. These kind of scenarios are complex and the data is scarce, particular when it comes to using infrared sensors. Tiltan comes in and fills the training gap that AI companies have in developing such AI models. In addition, Tiltan's 3D reconstruction solution (AGM) can assist first responders such as police when  a 3D representation of an area is required during operations for purposes such as access to a building (hostage scenario), analysis of communication coverage, etc. Tiltan invites all Media members to visit Booth #1619, 13-14.6.25 at 16:00 for a 1:1 briefing and demo.                                                                                                                                                                                             

  • Ultralytics (Booth 1512) - Ultralytics is the leading force in Vision AI, known for its revolutionary YOLO (You Only Look Once) models. With over 110K GitHub stars, 95+ million Pip downloads, 5 Million monthly visits, and 1.4B images analyzed per day, Ultralytics YOLOv5, YOLOv8, and YOLO11 have become the most widely recognized object detection models globally. Ultralytics empowers users with easy-to-use, cutting-edge AI technology. Our mission is to simplify the use of AI, making it accessible and impactful across industries ranging from manufacturing to healthcare and more.

  • Voxel51 (Booth 1417) - The visual AI data platform is highlighting its groundbreaking research showing that zero-shot auto-labeling can achieve accuracy nearly equivalent to human labeling (up to 95%) while operating 5,000x faster than traditional annotation methods. See a demo of how FiftyOne Verified Auto Labeling can reduce labeling costs by up to 100,000x—potentially saving millions of dollars in AI development costs. Visit the booth and check out our talks, tutorials and joint demos - including those with Nvidia Omniverse re: smart simulation, curation, visualization, and Databricks + Twelve Labs re: video search.

Advance News

  • 3dMD (Booth 1529) - Breaking the Barriers of Computing Human-to-Human Interactions. 3dMD announces its latest dynamic-3D/4D 3dMDbody System that captures Real-World 4D datasets of more than one person to help Computer-Vision scientists understand and model complex human-to-human interactions and natural behaviors. Implementing 25 years of medical-grade 3D imaging technology within a flexible modular configuration, 3dMD records complex nuanced gestures, expressions, gait, speech (with audio-synch), and contact between subjects engaged in interaction and communication. 3dMD subsequently outputs progressive sequences of anatomically-accurate 3D images/models of humans that digitally document and quantify the 360-degree spatial relationships and correlations of natural human interactions in real life scenarios. Click for demo. For more information: https://3dmd.com/ | info@3dMD.com | https://vimeo.com/3dmd

  • Abaka AI (Booth 1535) - Abaka AI is proud to be a sponsor at CVPR 2025! Visit us at Booth #1535 to explore how our end-to-end data solutions spanning data collection, cleaning, annotation, and high-quality dataset delivery, are empowering industry leaders in Generative AI, Autonomous Driving, and Embodied Agents. Our team specializes in delivering scalable, expert-labeled datasets tailored to complex AI challenges across different modalities. Whether you need finely tuned benchmarks or any large-scale training data, Abaka AI is your trusted partner for production-grade AI development! Let's connect and push the boundaries of computer vision together: contact@abaka.ai.

  • aiMotive (Booth 931) - aiMotive brings simulation closer to reality with the world’s first neural-reconstructed automotive proving ground, developed in collaboration with AVL ZalaZONE. This breakthrough demonstrates how real-world test environments can be digitally recreated with centimeter-level accuracy and photorealistic rendering. Uniquely, the pipeline supports all key sensor modalities — LiDAR, radar, and camera — enabling complete scenario-based testing in simulation. This marks a major step toward bridging the domain gap between synthetic and real-world validation for ADAS and autonomous driving systems. For more details, read our article here.

  • Tiltan (Booth 1619) - Tiltan Software Engineering Ltd., a global leader in simulation, 3D engines and geo-mapping systems, is set to showcase its latest range of computer vision solutions to the upcoming conference in Tennessee. At the forefront of Tiltan’s portfolio is Majestic.ai, their flagship, multi-sensor simulation solution that delivers ultra-realistic “synthetic data” to accelerate and enhance AI/ML model training and validation. For more details, read our press release here. 

  • Vision Components (Booth 1129) - Plug&play Embedded Vision with VC MIPI Bricks. Vision Components will showcase its new VC MIPI Bricks system at the Computer Vision and Pattern Recognition Conference (CVPR), held from June 11-15 in Nashville, Tennessee. The modular VC MIPI Bricks system comprises perfectly matching camera modules, accessories and services, right through to ready-to-use MIPI cameras and complete embedded vision systems. A development kit from PHYTEC with NXP i.MX 8M Plus and i.MX 8M Mini processor is also part of the VC MIPI Bricks system. It works plug-and-play with all 50+ MIPI cameras from Vision Components and will be presented with a demo at the booth. The kit is immediately available and can be ordered from Vision Components and PHYTEC. For more details, read our press release here. 

  • Voxel51 (Booth 1417) - Visual AI data platform Voxel51 will issue at CVPR groundbreaking research showing that zero-shot auto-labeling can achieve accuracy nearly equivalent to human labeling (up to 95%) while operating 5,000x faster than traditional annotation methods. Labeling costs can be reduced by up to 100,000x—potentially saving millions of dollars in AI development costs. The findings were used in the development of Verified Auto Labeling, FiftyOne’s AI-assisted annotation solution that will combine automated labeling with targeted human QA workflows. To learn more, visit www.voxel51.com or visit us at CVPR booth #1417.

Other News