Gen AI_Object Automation_TechnologyWorkshop
- 1. Face Recognition Generative AI Using Azure
- 2. Introduction Harnessing the power of AI using Microsoft Azure • Definition of Face Recognition • Importance in Various Fields (Security, Marketing, Healthcare, etc • Potential Applications Scientific findings 2
- 3. Definition of Generative AI • Gen AI possess advanced capabilities in learning, reasoning, and adaptation, often approaching or surpassing human-level intelligence across a broad range of tasks and domains • Power of autonomous decision-making compared to earlier generations of AI. • Gen AI represents a significant advancement in the field of artificial intelligence and holds the potential to revolutionize various industries and aspects of society. Scientific findings 3
- 4. How does Gen AI work? • Learn Pattern in a dataset • Training • Deep learning, Adversarial learning, Reinforcement learning • Deep learning is a subfield of machine learning that involves the training and usage of artificial neural networks to perform various tasks • Adversarial learning involves iterative optimization o • Reinforcement learning where an agent learns to make decisions by interacting with an environment with feedback and optimization 4
- 5. Models used in Gen AI Generative Model • try to understand the structure of the data and uses it to generate new data similar to the original data Discriminative models • focus on the differences between the data • boundary that separates the different categories of data • eg: image generation, text generation, and even generating realistic-sounding speech 5
- 6. Discriminative Vs Generative Scientific findings 6
- 7. GAN • A generative adversarial network (GAN) is a deep learning architecture. It trains two neural networks to compete against each other to generate more authentic new data Style GAN DALL-E 7
- 8. Transformer Model (processes i/p sq) (gen o/p seq)) Bidirectional Encoder Representations from Transformers Generative Pretrained Transformers 8
- 9. Uses of Generative models • Text Generation • Sentiment Analysis • Image Generation and Enhancement • Video creation • Code Generation • Speech to Speech conversion (STS) 9
- 10. • Text-to-Speech generation (TTS) • Audio generation • Synthetic data generation and augmentation Scientific findings 10
- 11. Use cases of generative AI models across domains Use cases of Generative AI models across domains • Healthcare • Art and Animation • Marketing and Sales • Software Programming • Finance • Manufacturing • Entertainment Scientific findings 11
- 12. Prominent examples of Generative AI tools • Chat GPT • GitHub Copilot • Pictory.AI • Midjourney • Wordtune • Gretel • Genie AI Scientific findings 12
- 13. Azure AI Services Introduction to Azure AI Services Overview of Azure Cognitive Services Importance of Azure for AI Development Azure Face API Azure Custom Vision Azure Machine Learning Scientific findings 13
- 14. Azure Face API • Used in security systems, access control, user authentication, personalized experiences, and content moderation. • Face Detection • Face Verification • Face Identification • Face Recognition Scientific findings 14
- 15. Azure Custom Vision • Pre-built AI models that provide capabilities for vision, speech, language • Computer Vision: extract information from images and videos, including image analysis, object detection, image classification, and optical character recognition(OCR). • Language Services: Offers natural language processing (NLP) capabilities, including text analysis, sentiment analysis, language translation Scientific findings 15
- 16. Azure Machine Learning • cloud-based platform that enables data scientists and machine learning practitioners to build, train, deploy, and manage machine learning models at scale • offers a wide range of tools and capabilities to streamline the end-to-end machine learning workflow, from data preparation to model deployment. Scientific findings 16
- 17. Face Recognition Generative AI • The integration of generative models with face recognition systems can lead to more robust, accurate, and versatile systems capable of performing well in a wide range of scenarios and conditions Scientific findings 17
- 18. Steps to create a face recognition system using Generative AI • Data Collection and Preparation • Train a Generative Model(GAN) • Train a Face Recognition Model • Choose a suitable face recognition algorithm or model architecture, such as a convolutional neural network (CNN) trained on embeddings of facial features. • Train the face recognition model to learn discriminative features that distinguish between different individuals in the dataset. • Integration of Generative Model with Face Recognition 18
- 19. Facial Recognition System 19 • Register User, Train Image, Trigger are simple Azure functions ∙ Training Faces of Individuals ∙ Identifying faces of Individuals
- 20. Case Studies • Real-world Examples of Face Recognition Generative AI Projects • Virtual Try-On in Fashion, Facial Expression Generation, Personalized Avatar Creation, Facial Aging Simulation Scientific findings 20
- 21. Ethical Considerations • Ethical Implications of Face Recognition Technology • Privacy Concerns and Data Security • Accuracy and Misidentification Scientific findings 21
- 22. Future Directions • Biometric Face Recognition for fraud prevention • Character Analysis • Opportunities for Further Research and Innovation Scientific findings 22
- 23. Summary of Key Points • Importance of Azure in Advancing Face Recognition Generative AI • Questions and Discussion Scientific findings 23