# DTBS: Dual-Teacher Bi-directional Self-training for Domain Adaptation in Nighttime Semantic Segmentation

arXiv: https://arxiv.org/abs/2401.01066 · HTML: https://mlsanity.com/paper/2401.01066/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 7 of its 40 references that are in the map.
- First posted on arXiv 2 Jan 2024.

## Builds on (most cited first, 7 of 7)

- [SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers](https://mlsanity.com/paper/2105.15203/) (2021)
- [Decoupled Weight Decay Regularization](https://mlsanity.com/paper/1711.05101/) (2017)
- [Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time](https://mlsanity.com/paper/2203.05482/) (2022)
- [Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results](https://mlsanity.com/paper/1703.01780/) (2017)
- [Debiased Self-Training for Semi-Supervised Learning](https://mlsanity.com/paper/2202.07136/) (2022)
- [Deliberated Domain Bridging for Domain Adaptive Semantic Segmentation](https://mlsanity.com/paper/2209.07695/) (2022)
- [Guided Curriculum Model Adaptation and Uncertainty-Aware Evaluation for Semantic Nighttime Image Segmentation](https://mlsanity.com/paper/1901.05946/) (2019)

## Built on by (most cited first, 4 of 4)

- [SCRIBE: Structured Mid-Level Supervision for Tool-Using Language Models](https://mlsanity.com/paper/2601.03555/) (2026)
- [COSMIC: Clique-Oriented Semantic Multi-space Integration for Robust CLIP Test-Time Adaptation](https://mlsanity.com/paper/2503.23388/) (2025)
- [Exposure Bias Can Alleviate Itself via Directional and Frequency Rectification in Flow Matching](https://mlsanity.com/paper/2606.28226/) (2026)
- [Decoupled Competitive Framework for Semi-supervised Medical Image Segmentation](https://mlsanity.com/paper/2505.24667/) (2025)

## Abstract (arXiv)

Due to the poor illumination and the difficulty in annotating, nighttime conditions pose a significant challenge for autonomous vehicle perception systems. Unsupervised domain adaptation (UDA) has been widely applied to semantic segmentation on such images to adapt models from normal conditions to target nighttime-condition domains. Self-training (ST) is a paradigm in UDA, where a momentum teacher is utilized for pseudo-label prediction, but a confirmation bias issue exists. Because the one-directional knowledge transfer from a single teacher is insufficient to adapt to a large domain shift. To mitigate this issue, we propose to alleviate domain gap by incrementally considering style influence and illumination change. Therefore, we introduce a one-stage Dual-Teacher Bi-directional Self-training (DTBS) framework for smooth knowledge transfer and feedback. Based on two teacher models, we present a novel pipeline to respectively decouple style and illumination shift. In addition, we propose a new Re-weight exponential moving average (EMA) to merge the knowledge of style and illumination factors, and provide feedback to the student model. In this way, our method can be embedded in other UDA methods to enhance their performance. For example, the Cityscapes to ACDC night task yielded 53.8 mIoU (%), which corresponds to an improvement of +5% over the previous state-of-the-art. The code is available at urlhttps://github.com/hf618/DTBS.