The paper titled ’A Novel Non-iterative Training Method for CNN Classifiers Using Gram-Schmidt Process’ has been accepted for publication in Neural Processing Letters 2025. 

The paper titled "A Novel Graph-Based Framework for Understanding Decision-Making in Deep Learning Models" has been accepted in the IEEE Symposium Series on Computational Intelligence (SSCI), 2025, Trondheim, Norway. 

The paper titled "Explainable Image Recognition with Graph-based Feature Extraction" has been accepted in IEEE Access Journal 2024.

Summary

This ARC Discovery project aims to study feature extraction abilities of convolutional as well as traditional neural networks and develop a generic feature extractor which can be applied to wide variety of real-world image and non-image data. New concepts for automatic feature extraction, feature explanation, hybrid evolutionary algorithms and non-iterative ensemble learning will be introduced and evaluated. The expected outcomes are a generic feature extractor for automatically extracting features, an optimizer for finding optimal parameters and non-iterative ensemble learning technique for classification of features into classes. The impact of this project will be automatic feature extractors and classifiers for real-world applications. 

Code and Data

The codes and data being developed for the project are available for each manuscript alongside an author version of the paper below. 

 Presentations

• SSCI - 2025: A Novel Graph-based Framework for Understanding of Decision-Making Process in Deep Learning Models. [Slides] [Yet to be Presented]

• ICONIP - 2024: Optimizing CNNs with Gram Schmidt Non-Iterative Learning for Image Recognition. [Slides]

• WCCI - 2024: Neuron Efficiency Index: An Empirical Method for Optimizing Parameters in Deep Learning [Slides]

• DICTA-2023: A Novel Graph-based Framework for Explainable Image Classification: Features That Matter. [Slides]

• IJCNN-2023: Novel Automatic Deep Learning Feature Extractor with Target Class Specific Feature Explanations. [Slides]

• DICTA-2022:  Feature Extractor Based on Class Specific Hidden Neuron Activations for Image Classification. [Slides]

• IVCNZ-2022:  A Novel Explainable Deep Learning Model with Class Specific Features. [Slides]

• The slides presenting the progress of the project can be found here. [Slides]

 Research Publications

   The journal and research papers produced under this project are:

1. Kuttichira DP, Verma B, Azam B, Rahman A, Wang L. Feature Extractor Based on Class Specific Hidden Neuron Activations for Image Classification. In 2022 International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2022 Nov 30 (pp. 1- 6). IEEE.  [PDF][Online][Code]

2. Kuttichira DP, Azam B, Verma B, Rahman A, Wang L. A Novel Explainable Deep Learning Model with Class Specific Features. In International Conference on Image and Vision Computing New Zealand 2022 Nov 23 (pp. 62-74). Cham: Springer Nature Switzerland.[PDF][Online][Code]

3. Kuttichira DP, Azam B, Verma B, Rahman A, Wang L. Novel Automatic Deep Learning Feature Extractor with Target Class Specific Feature Explanations. In International Joint Conference on Neural Networks (IJCNN) 2023 Jun 18 (pp. 1-8). [PDF][Online][Code]

4. Kuttichira DP, Azam B, Verma B, Rahman A, Wang L, Abdul Sattar. Neural Network Feature Explanation Using Neuron Activation Rate Based Bipartite Graph. In International Conference on Image and Vision Computing New (IVCNZ) 2023 Nov 29 (pp. 1-6). [PDF] [Online][Code]

5. Azam B, Kuttichira DP, Verma B, Rahman A, Wang L, Abdul Sattar. A Novel Graph-based Framework for Explainable Image Classification: Features That Matter. In International Conference on Digital Image Computing: Techniques and Applications (DICTA) 2023. [PDF] [Online] [Code]

6. Azam B, Kuttichira DP, and Verma B. "Neuron Efficiency Index: An Empirical Method for Optimizing Parameters in Deep Learning" in The IEEE World Congress on Computational Intelligence (IEEE WCCI) 2024.  [PDF] [Online] [Code]

7. Kuttichira DP, Azam B, Verma B. , Rahman A, and Wang L."Optimizing CNNs with Gram Schmidt Non-Iterative Learning for Image Recognition" in International Conference in Neural Information Processing (IEEE ICONIP) 2024.  [PDF] [Will be Available Online April 2025][Code]

8. Azam, B. Kuttichira DP, B. Verma, A Rahman, and L. Wang "Explainable Image Recognition with Graph-based Feature Extraction " in IEEE ACCESS 2024.  [PDF] [Online]

9. Azam B., P. Sanjeewani, B. Verma, A Rahma, and L. Wang "A Novel Graph-Based Framework for Understanding of Decision-Making Process in Deep Learning Models" in IEEE Symposium Series on Computational Intelligence (SSCI) 2025.  [PDF] [Will be Available Online June 2025]

10. Azam, B. D. Kuttichira P. Sanjeewani, B. Verma, A Rahma, and L. Wang "A Novel Non-iterative Training Method for CNN Classifiers Using Gram-Schmidt Process" in Neural Processing Letters, 2025.  [PDF]