Fetching the paper…

Training Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomization · Around