Publications

📄 Peer-Reviewed Publications

• Ch M Awais, M. Reggiannini, D. Moroni
  A Framework for Imbalanced SAR Ship Classification: Curriculum Learning, Weighted Loss Functions, and a Novel Evaluation Metric
  WACV 2025, Tucson, Arizona, USA.
  [Paper Link]
  Abstract

  Introduces a curriculum learning framework, weighted losses, and a novel evaluation metric to improve SAR ship classification on imbalanced datasets.
• Ch M Awais, M. Reggiannini, D. Moroni
  Image Quality vs Performance in Super-Resolution for SAR-Ship Classification
  ISCAS 2025, London, UK.
  [Paper Link]
  Abstract

  Explores the relationship between super-resolved image quality and classification performance on SAR ship datasets using deep learning models.
• Ch M Awais, M. Reggiannini, D. Moroni, O. Karakus
  Feature-Space Oversampling for Addressing Class Imbalance in SAR Ship Classification
  IGARSS 2025, Brisbane, Australia (Accepted).
  [Paper Info]
  Abstract

  Applies feature-space augmentation to rebalance data distribution in SAR ship classification, improving performance on underrepresented classes.
• Ch M Awais, M. Reggiannini, D. Moroni, A. Galdelli
  SAR-to-Infrared Domain Adaptation for Maritime Surveillance with Limited Data
  AITA 2025, Kobe, Japan (Accepted).
  [Paper Info]
  Abstract

  Introduces domain adaptation between SAR and IR imagery to improve maritime surveillance systems with limited labeled IR data.
• Ch M Awais, M. Reggiannini, D. Moroni
  Deep Learning for SAR Ship Classification: Focus on Unbalanced Datasets
  ICEAA-IEEE APWC 2024, Lisbon, Portugal.
  [Paper Link]
  Abstract

  Investigates the effect of class imbalance in SAR ship datasets and evaluates mitigation strategies using deep learning pipelines.
• Ch M Awais, I.E.I. Bekkouch, A.M. Khan
  What Augmentations Are Sensitive to Hyper-Parameters and Why?
  SAI 2022: Intelligent Computing
  [Springer Link]
  Abstract

  Empirically examines the sensitivity of common image augmentations to hyperparameter settings in classification tasks.
• Ch M Awais, W. Gu, G. Dlamini, Z. Kholmatova, G. Succi
  A Meta-Analytical Comparison of Naive Bayes and Random Forest for Software Defect Prediction
  ISDA 2022
  [Springer Link]
  Abstract

  Performs a comparative meta-analysis of Naive Bayes and Random Forest on defect prediction benchmarks across multiple datasets.

🖼️ Posters

• Testing a SAR-based ship classifier with different loss functions
  Presented at Biodiversity change in the Anthrospace: Priorities for research
  [View Poster PDF]
• Evaluating the Impact of Fine-tuning on Deep Learning Models for SAR Ship Classification
  Presented at Convegno Scientifico, 20-21 May 2024, Università degli Studi di Palermo
  [View Poster PDF]
• FUTURE FISHERY & DEEP LEARNING
  Presented at ISIT-DAY, Pisa, 2024
  [View Poster PDF]
• Novel Loss Function based Super-Resolution for SAR Ship Classification
  Presented at IGARSS, 2025
  [View Poster Info]