DeepScanner: a Robotic System for Automated 2D Object Dataset Collection with Annotations
This addresses the bottleneck of time-consuming and error-prone manual annotation for computer vision researchers and practitioners, representing a strong incremental improvement in dataset creation efficiency.
The paper tackles the problem of manual dataset labeling for 2D objects by introducing an automated robotic system, resulting in a 240-fold increase in labeling speed and a 13-fold improvement in accuracy compared to manual methods.
In the proposed study, we describe the possibility of automated dataset collection using an articulated robot. The proposed technology reduces the number of pixel errors on a polygonal dataset and the time spent on manual labeling of 2D objects. The paper describes a novel automatic dataset collection and annotation system, and compares the results of automated and manual dataset labeling. Our approach increases the speed of data labeling 240-fold, and improves the accuracy compared to manual labeling 13-fold. We also present a comparison of metrics for training a neural network on a manually annotated and an automatically collected dataset.