Plainsight Pipelines, the latest addition to our On-Demand computer vision platform, makes it even easier for users to build and scale computer vision models, prediction pipelines, and custom data transformations.
What Is Plainsight Pipelines?
Pipelines is Plainsight’s interface for automating and streamlining the process of feeding visual data into, out of, and through our computer vision platform. The interface provides an easy method for users to upload visual data from cloud buckets or streaming edge nodes, enriching unstructured data, and deploying computer vision models tailored to address their specific use cases.
- Streaming and Batch Inputs: Pipelines supports streaming (always on) and batch (one-off or scheduled) ingestions of visual data via cloud buckets, Google Pub/Sub, and/or API endpoints.
- Modular Pipeline Blocks: Pipelines allows for sequential processing of the visual data users upload. Users can set up automated processing augmentations like resizing, cropping, and tiling to manage and transform data at scale.
- Pre-Built and Custom Model Blocks: Users can deploy their own Plainsight-trained models into their pipelines or use our off-the-shelf Object Detection model, capable of detecting more than 80 types of objects.
Why Plainsight Pipelines?
Plainsight Pipelines guides users through the process of creating and managing visual data workflows. It eliminates the need for tedious, manual data handling as well as complex scripting tools and even specialized domain experience. Users of all experience levels can save time and effort with Plainsight Pipelines and devote their attention to more high-impact objectives. What’s more, the modular design of Pipeline Blocks speeds innovation, enabling users to experiment with ease.
Get Started with Pipelines:
- Sign-Up Plainsight On-Demand or log in your existing account.
- Once into your account, make your way over to feature and create your first Pipeline!
- For more information on the Pipelines feature, visit our docs.
- Send your questions and feedback to email@example.com or reach out via our community Slack channel to share your thoughts with our team.