One paper has been accepted to ICLR 2023. ICLR is one of the major international conferences on machine learning and related areas.
Title: DFlow: Learning to Synthesize Better Optical Flow Datasets via a Differentiable Pipeline
Authors: Kwon Byung-Ki (POSTECH), Nam Hyeon-Woo (POSTECH), Ji-Yun Kim (Krafton), Tae-Hyun Oh (POSTECH)
* This work was done when Ji-Yun Kim was in POSTECH.
[Abstract]
Comprehensive studies of synthetic optical flow datasets have attempted to reveal
what properties lead to accuracy improvement. However, manually identifying
and verifying all such necessary properties are intractable mainly due to the requirement of large-scale trial-and-error experiments with iteratively generating whole synthetic datasets. To tackle this challenge, we propose a differentiable
optical flow data generation pipeline and a loss function to drive the pipeline called DFlow. These enable automatic and efficient synthesis of a dataset effective to a target domain. This distinctiveness is achieved by proposing an efficient data comparison method, where we approximately encode reference sets of data into neural networks and compare the proxy networks instead of explicitly comparing datasets in a sample-wise way. Our experiments show the competitive performance of our DFlow against the prior arts in pre-training.
[Results]
