For companies:
For professionals:
For everyone:
Responsibilities
-Implement, train and iterate vision based deep learning models.
-Train models on Cloud on single and multiple GPUs.
-Design and implement data pipelines for storing and processing thousands of images and videos.
-Design and integrate with third party labeling company API for upload and ingestion of data and labels.
-Design and implementation of data harvesting from current fleet operations.
-Create data visualization tools for data analysis and debugging.
-Evaluate the quality of labeling of data.
-Design and iterate labeling instructions that produce high quality labels.
-Implement, design and manage databases of images, videos and metadata associated with them.
-Optimize models for performance on constrained environments.
-Deploy trained models on-premise integrated with the rest of the robot’s software stack.
-Report of training procedures, error analysis, data review.
-Offline database analysis for getting insights, generating labels automatically, doing anonymization, etc.
This job post is closed.