Summary: Automatic assessment of road safety and conditions is essential for improving road infrastructure and reducing fatalities on the roads. The current manual systems used for road safety not only in Australia but around the world are inefficient and prone to many errors. The major challenges are to accurately detect, segment and classify all road objects and also calculate the distance between objects. Deep learning with a recent breakthrough has the ability to address such major challenges. The main aim of this project is to develop a novel deep learning based automated system that can analyze video data and assess road safety and conditions.   

Presentation and Demo: Poster - Image and Vision Computing New Zealand (IVCNZ) 2019 (Click Here)

Presentation - Image and Vision Computing New Zealand (IVCNZ) 2020 (Click Here)

Presentation - IEEE Congress on Evolutionary Computation (CEC) 2021 (Click Here)

Presentation - IEEE Symposium Series on Computational Intelligence (SSCI) 2021 (Click Here)

Poster - Image and Vision Computing New Zealand (IVCNZ) 2021 (Click Here)

Code and Data: The code and data for the developed deep learning based technique are available on GitHub and Google Drive. You can download Code (Click Here) and Data (Click Here). We encourage you to run the code with different data including your own data and let us know any new results you get. Any questions or feedback can be sent to Prof. Brijesh Verma by email (b.verma@cqu.edu.au or b.verma.qld@gmail.com). 

Social Media Posts: https://twitter.com/BrijeshVermaQld/status/1272774055330897920 

Programs and Systems: https://github.com/bvermaqld/DeepLearning 

Publications

1.  P. Sanjeewani and B. Verma, Single Class Detection-based Deep Learning Approach for Identification of Road Safety Attributes, Neural Computing and Applications, vol. 33, pp. 9691-9702, 2021. [Pdf][Available Online] 

2.  P. Sanjeewani and B. Verma, Optimization of Fully Convolutional Network for Road Safety Attribute Detection, IEEE Access, vol. 9, pp. 120525-120536, 2021. [Pdf][Available Online] 

3. P. Sanjeewani, B. Verma and J. Affum, A Novel Evolving Classifier with a False Alarm Class for Speed Limit Sign Recognition, IEEE Congress on Evolutionary Computation (CEC), Krakow, Poland, 2021. [Pdf][Available Online]

4. P. Sanjeewani, B. Verma and J. Affum, Multi-stage Deep Learning Technique with Cascaded Classifier for Improving Turn Lanes Recognition, IEEE Symposium Series on Computational Intelligence (SSCI), Orlando, Florida, USA, 2021. [Pdf][Available Online]

5. P. Sanjeewani, B. Verma and J. Affum, Multi-stage Deep Learning Technique for Improving Traffic Signs Recognition, 36th International Conference on Image and Vision Computing New Zealand (IVCNZ), Waikato, New Zealand, pp. 1-6, 2021. [Pdf][Available Online]

6. M. Zhong, B. Verma and J. Affum, Multi-Receptive Atrous Convolutional Network for Semantic Segmentation, International Joint Conference on Neural Networks (IJCNN), Glasgow, UK, pp. 1-8, 2020. [Pdf][Available Online]

7. P. Sanjeewani and B. Verma, An Optimization Technique for the Detection of Safety Attributes using Roadside Video Data, 35th International Conference on Image and Vision Computing New Zealand (IVCNZ), Wellington, New Zealand, pp. 1-6, 2020. [Pdf][Available Online]

8. M. Zhong, B. Verma and J. Affum, Deep 3D Segmentation and Classification of Point Clouds for Identifying AusRAP Attributes, International Conference on Neural Information Processing (ICONIP), Sydney, Australia, pp. 95-105, 2019. [Pdf][Available Online]

9. T. G. P. Sanjeewani and B. Verma, Learning and Analysis of AusRAP Attributes from Digital Video Recording for Road Safety, 34th International Conference on Image and Vision Computing New Zealand (IVCNZ), Dunedin, New Zealand, pp. 1-6, 2019. [Pdf][Available Online]

10. M. Zhong, B. Verma and J. Affum, Point Cloud Classification for Detecting Roadside Safety Attributes and Distances, IEEE Symposium Series on Computational Intelligence (SSCI), Xiamen, China, pp. 1078-1084, 2019. [Pdf][Available Online]

11. Z. Jan, B. Verma, J. Affum, S. Atabak and L. Moir, A Convolutional Neural Network based Deep Learning Technique for Identifying Road Attributes, International Conference on Image and Vision Computing New Zealand (IVCNZ), Auckland, New Zealand, pp. 1-6, 2018. [Pdf][Available Online]

12. A. G. Perera and B. Verma, Road Severity Distance Calculation Technique Using Deep Learning Predictions in 3-D Space, IEEE Access, vol. 10, pp. 68000 - 68008, 2022.  [Pdf][Available Online]