Learning Lightweight Lane Detection CNNs by Self Attention Distillation
Yuenan Hou,Zheng Ma,Chunxiao Liu,Chen Change Loy +3 more
- pp 1013-1021
TLDR
Self Attention Distillation (SAD) as discussed by the authors is a knowledge distillation approach, which allows a model to learn from itself and gains substantial improvement without any additional supervision or labels.Abstract:
Training deep models for lane detection is challenging due to the very subtle and sparse supervisory signals inherent in lane annotations. Without learning from much richer context, these models often fail in challenging scenarios, e.g., severe occlusion, ambiguous lanes, and poor lighting conditions. In this paper, we present a novel knowledge distillation approach, i.e., Self Attention Distillation (SAD), which allows a model to learn from itself and gains substantial improvement without any additional supervision or labels. Specifically, we observe that attention maps extracted from a model trained to a reasonable level would encode rich contextual information. The valuable contextual information can be used as a form of ‘free’ supervision for further representation learning through performing top- down and layer-wise attention distillation within the net- work itself. SAD can be easily incorporated in any feed- forward convolutional neural networks (CNN) and does not increase the inference time. We validate SAD on three popular lane detection benchmarks (TuSimple, CULane and BDD100K) using lightweight models such as ENet, ResNet- 18 and ResNet-34. The lightest model, ENet-SAD, performs comparatively or even surpasses existing algorithms. Notably, ENet-SAD has 20 × fewer parameters and runs 10 × faster compared to the state-of-the-art SCNN, while still achieving compelling performance in all benchmarks.read more
Citations
More filters
Journal ArticleDOI
Knowledge Distillation: A Survey
TL;DR: A comprehensive survey of knowledge distillation from the perspectives of knowledge categories, training schemes, teacher-student architecture, distillation algorithms, performance comparison and applications can be found in this paper.
Posted Content
Ultra Fast Structure-aware Deep Lane Detection
Zequn Qin,Huanyu Wang,Xi Li +2 more
TL;DR: A novel, simple, yet effective formulation aiming at extremely fast speed and challenging scenarios, which treats the process of lane detection as a row-based selecting problem using global features and proposes a structural loss to explicitly model the structure of lanes.
Journal ArticleDOI
Visual Perception Enabled Industry Intelligence: State of the Art, Challenges and Prospects
TL;DR: The previous research and application of visual perception in different industrial fields such as product surface defect detection, intelligent agricultural production, intelligent driving, image synthesis, and event reconstruction are reviewed.
Book ChapterDOI
Ultra Fast Structure-aware Deep Lane Detection
Zequn Qin,Huanyu Wang,Xi Li +2 more
TL;DR: In this paper, Liu et al. proposed a novel, simple, yet effective formulation aiming at extremely fast speed and challenging scenarios, which treated the process of lane detection as a row-based selecting problem using global features.
Posted Content
Key Points Estimation and Point Instance Segmentation Approach for Lane Detection.
TL;DR: A traffic line detection method called Point Instance Network (PINet), based on the key points estimation and instance segmentation approach, which achieves competitive accuracy and false positive on the TuSimple and Culane datasets.
References
More filters
Proceedings ArticleDOI
Deep Residual Learning for Image Recognition
TL;DR: In this article, the authors proposed a residual learning framework to ease the training of networks that are substantially deeper than those used previously, which won the 1st place on the ILSVRC 2015 classification task.
Posted Content
Distilling the Knowledge in a Neural Network
TL;DR: This work shows that it can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model and introduces a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse.
Proceedings Article
Spatial transformer networks
TL;DR: This work introduces a new learnable module, the Spatial Transformer, which explicitly allows the spatial manipulation of data within the network, and can be inserted into existing convolutional architectures, giving neural networks the ability to actively spatially transform feature maps.
Proceedings Article
Multi-Scale Context Aggregation by Dilated Convolutions
Fisher Yu,Vladlen Koltun +1 more
TL;DR: This work develops a new convolutional network module that is specifically designed for dense prediction, and shows that the presented context module increases the accuracy of state-of-the-art semantic segmentation systems.
Book ChapterDOI
Large-Scale Machine Learning with Stochastic Gradient Descent
TL;DR: A more precise analysis uncovers qualitatively different tradeoffs for the case of small-scale and large-scale learning problems.