Fine-tuning pre-trained neural network models has become a widely adopted approach across various domains. However, it can lead to the distortion of pre-trained feature extractors that already possess strong generalization capabilities. Mitigating feature distortion during adaptation to new target domains is crucial. Recent studies have shown promising results in handling feature distortion by aligning the head layer on in-distribution datasets before performing fine-tuning. Nonetheless, a significant limitation arises from the treatment of batch normalization layers during fine-tuning, leading to suboptimal performance. In this paper, we propose Domain-Aware Fine-Tuning (DAFT), a novel approach that incorporates batch normalization conversion and the integration of linear probing and fine-tuning. Our batch normalization conversion method effectively mitigates feature distortion by reducing modifications to the neural network during fine-tuning. Additionally, we introduce the integration of linear probing and fine-tuning to optimize the head layer with gradual adaptation of the feature extractor. By leveraging batch normalization layers and integrating linear probing and fine-tuning, our DAFT significantly mitigates feature distortion and achieves improved model performance on both in-distribution and out-of-distribution datasets. Extensive experiments demonstrate that our method outperforms other baseline methods, demonstrating its effectiveness in not only improving performance but also mitigating feature distortion.

Radar is a key component of autonomous driving. We can capture the target range and velocity using radar signal that is transmitted and reflected by a target. However, the noise floor increases when interference signals exist, which severely affects the detectability of target objects. For these reasons, many previous studies have proposed methods for cancelling interference or reconstructing original signals. However, little work has been done using deep learning algorithm on interference mitigation. In this paper, we propose a new method using deep learning. We improve the performance of the existing deep learning algorithm using attention mechanism. We applied our algorithm to the OFDM radar environment as well as the existing frequency modulated continuous wave(FMCW) radar. The experiment show our deep learning based method outperforms existing methods. The implementation of our paper is available at https://github.com/jwmun/Radar.