Invited Speaker

Assoc. Prof. Deepika Koundal, University of Eastern Finland, Finland

Biography: Dr. Deepika Koundal currently serves as a Senior Researcher at the University of Eastern Finland, an Adjunct Researcher at Ho Chi Minh City Open University in Vietnam, and a Senior Associate Professor (Sabbatical) at UPES, Dehradun, India. She is having a rich academic experience, specializing in artificial intelligence, computer vision, image processing, and deep learning. She holds a B.Tech, M.Tech, and Ph.D. in Computer Science and Engineering and has received several prestigious accolades, including the MSCA Seal of Excellence from the European Commission. Recognized as a top 2% researcher by Stanford University in 2023 and 2024. She has earned multiple research excellence awards from UPES and Chitkara University and has published numerous research articles, edited notable books, and holds several patents. Additionally, she contributes as a guest and associate editor for leading journals, including IEEE and Elsevier.

Speech Title: From Neurons to Foundation Models: Modern Deep Learning Architectures

Speech Abstract: Modern deep learning has evolved from biologically inspired artificial neurons to powerful foundation models capable of solving diverse tasks across computer vision, natural language processing, speech recognition, and multimodal learning. This paper reviews the progression of deep learning architectures, beginning with perceptrons and multilayer neural networks, followed by convolutional neural networks, recurrent neural networks, long short-term memory networks, generative adversarial networks, transformers, and large foundation models. It highlights the architectural innovations, training paradigms, and applications that have shaped this evolution while discussing current challenges, including computational cost, interpretability, data dependency, and ethical concerns, along with emerging research directions.