Innoviz and Vueron team up to automate lidar labeling and speed perception development

Innoviz and Vueron integrate lidar data into VueX to speed up perception development

Innoviz Technologies and Vueron Technology are partnering to bring InnovizTwo and InnovizSmart lidar data into Vueron's VueX AI development platform. The goal: shorten the cycle from raw sensor captures to trained, validated perception models with automated annotation, training, and validation in one cloud workflow.

What's new

• Direct upload of Innoviz lidar data to the cloud-based VueX platform.
• Automated annotation of lidar point clouds with bounding boxes and object classes, with options to fine-tune for higher accuracy where needed.
• Support expanded from InnovizOne to Innoviz's latest: InnovizTwo for automotive and InnovizSmart for smart infrastructure.
• End-to-end environment covering data storage, visualization, processing, labeling, training, validation, and deployment.

Why this matters for developers

• Less manual labeling work and more consistent annotations across large datasets.
• Tighter feedback loop: iterate on labels, retrain, and validate without hopping across fragmented tools.
• Cleaner handoff between data, ML, and validation teams with a shared platform.
• Path to production for both automotive and infrastructure use cases using the same pipeline.

Workflow at a glance

• Capture data with InnovizTwo (automotive) or InnovizSmart (infrastructure).
• Upload to VueX for automated point cloud annotation.
• Review and fine-tune labels where precision is critical.
• Train and validate perception models within the same environment.
• Export results or deploy models/datasets as needed.

Built on an existing partnership

Vueron's platform previously supported InnovizOne. With this update, VueX now supports InnovizTwo and InnovizSmart, extending coverage across automotive and smart infrastructure. The emphasis is on a single AI cloud workflow for data lifecycle management-storage, visualization, processing, labeling, training, validation, and deployment.

What they're saying

Omer Keilaf, CEO and co-founder of Innoviz: "By integrating our automotive-grade sensors with Vueron's powerful AI development tools, we're enabling our customers to accelerate their development timelines and potentially reduce the complexity of bringing perception systems to market."

Noah Jang, SVP of global business at Vueron: "The combination of Innoviz's proven sensor technology with our automated annotation and training capabilities creates a powerful toolset that addresses real pain points in perception development."

Practical considerations before you adopt

• Define your ontology early and keep it consistent. Changes later can ripple through models and metrics.
• Decide where human-in-the-loop review is essential (e.g., edge cases, safety-critical classes).
• Track dataset and model versions from day one. Label changes should tie back to training runs and validation reports.
• Benchmark automated labels against your current baselines to quantify quality gains.
• Plan for deployment targets (vehicle ECUs, roadside units, or cloud) to avoid rework at the end.
• Align on KPIs that matter for your application (precision/recall per class, latency budgets, scenario coverage).

Where this helps

• Automotive ADAS/AD perception teams aiming to shorten data-to-model cycles.
• Smart infrastructure projects that need scalable labeling and validation across diverse environments.
• Organizations standardizing on lidar-first stacks and looking for a unified toolchain.

If your team is building end-to-end AI pipelines and needs structured upskilling, explore the latest hands-on programs at Complete AI Training.