Voxel's Transformation: Enhancing In-house Labeling Operations for High-Quality Training Data

• Analytics & Modeling - Computer Vision Software
• Platform as a Service (PaaS) - Application Development Platforms

• Education
• Equipment & Machinery

• Quality Assurance

• Computer Vision
• Visual Quality Detection

• System Integration
• Training

Voxel, a company leveraging AI and computer vision to manage risk and operations, faced two significant challenges. Firstly, they needed to maintain high-quality training data for their computer vision system. Secondly, they sought to automate their labeling process for faster throughput while retaining their in-house annotation team. Voxel had already invested in an in-house annotation team of subject matter experts, but they were struggling with efficiency in their labeling operations. They had been using an open-source solution, Computer Vision Annotation Tool (CVAT), which was causing bottlenecks as they increased the volume of annotations needed for model training. From an operational perspective, Voxel found it difficult to efficiently collect data and insights on the data labeling process, leading to significant manual effort. The tool couldn’t effectively link data quality to individual annotators, making it hard to identify the cause of low-quality labels. On the engineering side, Voxel had to custom-build data pipelines for new customer projects, a process that took multiple engineers four weeks for each project.

Voxel is a company that leverages AI and computer vision to change how companies manage risk and operations. They enhance their customers' security cameras with real-time AI to detect hazards, risky activities, and operational inefficiencies. Their technology delivers insights that can prevent injuries before they happen and improve the safety culture in hazardous and dynamic environments. To develop a robust computer vision system, Voxel requires large amounts of high-quality training data. They have an in-house annotation team of subject-matter experts who are well-versed in handling Voxel's specific use case.

Voxel partnered with Scale Studio to overcome their challenges. Scale Studio was chosen for its comprehensive management features, easy integration of data pipelines, ecosystem of integrated ML tools, and its experience with processing billions of annotations. Scale Studio's features included training courses, benchmark tasks, and annotator metrics such as throughput, efficiency, and accuracy. The platform also offered APIs for easy integration of data pipelines and quick setup of labeling projects. Scale Studio's ecosystem of integrated ML tools, such as Nucleus for dataset curation and management, and Rapid for Scale-managed dataset annotation, were also beneficial. Scale Studio's customer success and engineering teams provided strong technical knowledge and responsiveness, partnering with Voxel to ensure all frames of complex variable-frame-rate (VFR) videos were extracted to maximize the accuracy of the annotations and the model.

• After partnering with Scale Studio, Voxel was able to onboard their 20+ subject-matter experts onto the platform. This gave Voxel's data operations managers visibility into their in-house labeling team with annotator metrics such as throughput, efficiency, and accuracy. Studio also made it easy to streamline a data labeling process with intuitive tools and standardized workflows. They could now forecast labeling capacity and plan to match variable labeling demands. This resulted in a more efficient operation, saving time for both the operations managers and the engineering team. The engineering team also found it easier to integrate multiple data pipelines into their operations, reducing manual work.

• Voxel's operations managers saved 20% of time each week using Studio.
• With Studio, the engineering team cut their time by 50% to kick off new projects.