Welcome to IEEE SPS GenAI 2026 Seasonal School !!!

About IEEE SPS GenAI 2026

Generative Artificial Intelligence (GenAI) encompasses a class of computational models capable of producing novel content across multiple modalities, including text, images, audio, video, and synthetic data. These systems acquire representational knowledge from extensive training datasets through advanced deep learning techniques and subsequently generate outputs that exhibit statistical and structural similarity to the learned data distributions. The operational workflow of generative AI can be conceptualized in three primary phases: model training, task-specific tuning, and content generation. During the training phase, large-scale “foundation models” are developed by exposing neural architectures to diverse and voluminous datasets. The tuning phase further adapts these models to specialized tasks through approaches such as fine-tuning or reinforcement learning from human feedback. In the final phase, the model produces outputs that are iteratively evaluated and refined, often supplemented by mechanisms such as Retrieval-Augmented Generation (RAG) to enhance factual grounding and reduce error propagation.

Contemporary generative models now support a wide range of applications, including natural language generation, creative image synthesis, video manipulation, and speech modeling. Additionally, generative AI has emerged as a critical tool for producing synthetic datasets for algorithm training, simulation, and scientific research, thereby augmenting scenarios in which real data are scarce, sensitive, or costly to obtain.

Objectives: This seasonal school aims to provide participants with hands-on experience, from the foundations, under the guidance of some of the leading experts in the fields of Generative AI, Deep Learning, Visual Signal Processing, and various applications. The objectives of the seasonal school are highlighted as follows,

• Building awareness towards Generative AI approaches and applications in Visual Biomedical and Biometric Signal Processing.
• Building technical capacity and communities of research students, educators, R&D, and Industry persons in the trending areas of GenAI.
• Providing tutorial and hands-on demo sessions, where the participants can experiment with concepts and methods.
• Facilitating the learning of participants through the internationally renowned experts in the domain.

IEEE SPS GenAI 2026 Topics to be Covered:
The proposed IEEE Seasonal School on Generative AI (GenAI) and its applications to Visual Biomedical and Biometric Signal Processing will cover various fundamental and advanced topics, providing participants with a comprehensive understanding. The following are the tentative topics:

Lecture(s):

• Fundamentals of Visual Signal Processing,
• Introduction to Generative AI,
• Historical and architectural evolution of Generative AI,
• Probabilistic Modeling & Latent Variable Methods,
• Variational Autoencoders (VAEs),
• Generative Adversarial Networks (GANs),
• Diffusion Models and Score-Based Generators,
• Transformer Architectures for Generation,
• Foundation Models & Reinforcement Learning from Human Feedback (RLHF),
• Multimodal Generative Models,
• Image Enhancement & Restoration Using GenAI,
• Neural Rendering & Advanced Visual Generators,
• Evaluation Metrics & Benchmarking,
• Ethics, Governance & Responsible AI.

Hands-On:

• Introduction to Python, PyTorch, and Hugging Face tools
• Building a basic autoencoder for image reconstruction & implementing a simple GAN
• Fine-tuning a pre-trained diffusion model (Stable Diffusion)
• Prompt engineering & running a multimodal model for image captioning
• Image inpainting or restoration using diffusion models

IEEE SPS GenAI 2026 Attractions:

• Keynotes, lectures and demo-sessions by world-renowned international and national domain experts.
• Comprehensive coverage of Generative AI, deep learning and applications, and biometric solutions.
• Sponsored by IEEE Signal Processing Society (IEEE SPS).
• Coffee break and Banquet access to registered and confirmed participants in offline mode at IIIT Allahabad for the seasonal school.