# KDTwin: Task-Aware Knowledge Distillation for Lightweight Multi-Task Driving Scene Segmentation

arXiv: https://arxiv.org/abs/2609.18955 · HTML: https://mlsanity.com/paper/2609.18955/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Detection and dense prediction distillation, one of 12 directions in Knowledge distillation.
- Builds on 12 of its 16 references that are in the map.
- First posted on arXiv 28 Jul 2026.

## Placed in

- Knowledge distillation / [Detection and dense prediction distillation](https://mlsanity.com/distill/detection-and-dense-prediction-distillation/)

## Builds on (most cited first, 12 of 12)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning](https://mlsanity.com/paper/1805.04687/) (2018)
- [A White Paper on Neural Network Quantization](https://mlsanity.com/paper/2106.08295/) (2021)
- [A Comprehensive Overhaul of Feature Distillation](https://mlsanity.com/paper/1904.01866/) (2019)
- [Structured Knowledge Distillation for Dense Prediction](https://mlsanity.com/paper/1903.04197/) (2019)
- [Tversky loss function for image segmentation using 3D fully convolutional deep networks](https://mlsanity.com/paper/1706.05721/) (2017)
- [A Survey on Deep Neural Network Pruning-Taxonomy, Comparison, Analysis, and Recommendations](https://mlsanity.com/paper/2308.06767/) (2023)
- [BPKD: Boundary Privileged Knowledge Distillation For Semantic Segmentation](https://mlsanity.com/paper/2306.08075/) (2023)
- [TwinLiteNet: An Efficient and Lightweight Model for Driveable Area and Lane Segmentation in Self-Driving Cars](https://mlsanity.com/paper/2307.10705/) (2023)
- [TwinLiteNet+: An Enhanced Multi-Task Segmentation Model for Autonomous Driving](https://mlsanity.com/paper/2403.16958/) (2024)
- [MAPS-X: Explainable Multi-Robot Motion Planning via Segmentation](https://mlsanity.com/paper/2010.16106/) (2020)
- [TwinMixing: A Shuffle-Aware Feature Interaction Model for Multi-Task Segmentation](https://mlsanity.com/paper/2603.28233/) (2026)

## Abstract (arXiv)

Efficient perception models are essential for real-time autonomous driving, where accuracy and computational cost must be carefully balanced. However, applying knowledge distillation to multi-task driving scene segmentation is challenging because drivable-area and lane segmentation exhibit different spatial characteristics and class imbalance. We propose KDTwin, a task-aware distillation framework for lightweight multi-task segmentation networks. The proposed method performs distillation at both the shared encoder and task-specific decoders. Encoder-level pairwise distillation transfers spatial relational knowledge to enhance the student's shared representation. For the decoders, we use a weighted loss for drivable-area segmentation and a boundary-aware loss for lane segmentation, enabling task-adaptive knowledge transfer without increasing inference complexity. Experiments on BDD100K show consistent improvements across the evaluated CNN-based and Transformer-based student models without increasing inference-time parameters or FLOPs. The results show that designing distillation objectives according to task-specific characteristics can effectively enhance multi-task segmentation performance for autonomous driving. The source code is available at https://github.com/chequanghuy/KDTwin.