Leila Malihi is a researcher in cognitive science with specialization in computer vision, machine learning, and biomedical image analysis. Her work focuses on developing efficient and controllable deep learning frameworks, particularly model compression techniques such as sequential knowledge distillation and pruning, enabling deployment of high-performance neural networks on edge and resource-limited devices. She has contributed significantly to advancing automated medical image analysis, including wound classification, child face recognition, malaria parasite detection, cancer diagnosis, and ECG signal processing. Her research integrates convolutional neural networks, sparse coding, autoencoders, transfer learning, GAN-based synthetic data generation, and modern pattern-recognition techniques to build interpretable, scalable, and real-time AI systems. She has also explored neural network eigenspaces, principal eigenfeatures, and logistic regression probes to better understand the inner inference behavior of deep models. Leila’s scholarly output reflects her interdisciplinary approach, contributing to journals and international conferences in machine learning, medical informatics, and image processing. Her published work has received 88 Scopus citations from 85 documents, with 10 indexed documents and an h-index of 5, demonstrating a growing impact in the field. On Google Scholar, her research has accumulated 134 citations, with an h-index of 6 and an i10-index of 5, further highlighting the relevance of her contributions to computational healthcare, interpretable AI, and efficient deep learning architectures. Her profile reflects a strong commitment to bridging core AI innovation with real-world biomedical applications.
Profile
Scopus | Google Scholar
Featured Publications
Malihi, L., & Heidemann, G. (2023). Efficient and controllable model compression through sequential knowledge distillation and pruning. Journal of Big Data and Cognitive Computing.
Richter, M. L., Malihi, L., Windler, A. K. P., & Krumnack, U. (2023). Analyzing the inference process in deep convolutional neural networks using principal eigenfeatures, saturation, and logistic regression probes. Journal of Applied Research in Electrical Engineering.
Malihi, L., & Malihi, R. (2020). Single stuck-at faults detection using test generation vector and deep stacked sparse autoencoder. SN Applied Sciences, 2(10), 1–10.
Malihi, L., Ansari-Asl, K., & Behbahani, A. (2015). Improvement in classification accuracy rate using multiple classifier fusion toward computer vision detection of malaria parasite. Jundishapur Journal of Health Sciences, 7(3), 26–32.
Malihi, L., Ansari-Asl, K., & Behbahani, A. (2015). Computer-aided diagnosis of malaria parasite using pattern recognition methods. AJUMS Journals, 14(1), 65–74.