BROWSE
Detail
Conference Paper
Context-aware Image-to-Image Translation for Autonomous Driving
Cited 0 time in 
Share 
Share
- Authors
- Narae Yi, Minho Park, Dong-oh Kang
- Issue Date
- 2023-10
- Citation
- International Conference on Consumer Electronics (ICCE) 2023 : Asia, pp.130-132
- Publisher
- IEEE
- Language
- English
- Type
- Conference Paper
- Abstract
- The fusion of deep learning and autonomous driving presents challenges in computer vision, particularly in image-to-image translation to simulate real driving conditions. A notable constraint in the industry is the scarcity of diverse, realistic training data, hindering progress in capturing fine-grained details like weather changes. To address this, extensive research has explored deep learning-based image-to-image translation. Our work targets a critical issue within autonomous driving datasets: the loss of important objects in image translation. We introduce context-aware image-to-image translation (CAIT), embedding image context information by utilizing features from a pre-trained segmentation map prediction network.
- KSP Keywords
- Computer Vision(CV), Context Information, Context aware, Driving conditions, Image Context, Learning-based, MAP prediction, Prediction network, Segmentation map, autonomous driving, deep learning(DL)