![Literature survey
• The release of ChatGPT in November 2022, a conversation-focused
model that follows human instructions, further underscored the
feasibility of AGI in practical applications (Liu et al., 2023a). This
development has had a wide-ranging impact across various sectors,
including journalism (Liu et al., 2023c), education (Zhai, 2023; Liu
et al., 2023b), healthcare (Li et al., 2023; Liu et al., [n. d.]; Holmes
et al., 2023), industry (Dou et al., 2023), agriculture (Rezayi
et al., 2023), law (Bubeck et al., 2023), gaming (Bubeck et al., 2023),
and finance (Wu et al., 2023c), catalyzing a popular wave in AI (Liu
et al., 2023a, g, h).
• Rishi Bommasani, Drew A Hudson, Ehsan Adeli, Russ Altman, Simran
Arora, Sydney von Arx, Michael S Bernstein, Jeannette Bohg,
Antoine Bosselut, Emma Brunskill, et al. 2021.On the opportunities
and risks of foundation models.arXiv preprint
arXiv:2108.07258 (2021).](https://image.slidesharecdn.com/technicalseminar-240404004355-1dd56e1c/85/technical-seminar-pptx-on-multi-model-of-AI-7-320.jpg)
technical seminar.pptx on multi model of AI
- 1. VISVESVARAYA TECHNOLOGICAL UNIVERSITY "Jnana Sangama", Belgaum: 590 018 H.K.E Society’s SIR M VISVESVARAYA COLLEGE OF ENGINEERING (Affiliated to VTU - Belagavi, Approved by AICTE, Accredited by NAAC) Yeramarus Camp, Raichur-584135, Karnataka 2023-2024 TECHNICAL SEMINAR PRESENTATION ON “MULTIMODAL AI ” UNDER THE GUIDENCE OF DR.SHARAN KUMAR DEPARTMENT OF ELECTRONICS AND COMMUNICATION ENGINEERING
- 2. POWERING THE NEXT CHAPTER IN GENERATIVE AI MULTIMODAL AI
- 4. CONTENTS • Introduction • Literature survey • Block diagram • Applications • Future scope • Benefits and challenges • Conclusion • Reference
- 6. Introduction • Multi modal AI is an advanced form of artificial intelligence that is able to analyze and interpret multiple modes of data simultaneously allowing it to generate more accurate and human like responses.
- 7. Literature survey • The release of ChatGPT in November 2022, a conversation-focused model that follows human instructions, further underscored the feasibility of AGI in practical applications (Liu et al., 2023a). This development has had a wide-ranging impact across various sectors, including journalism (Liu et al., 2023c), education (Zhai, 2023; Liu et al., 2023b), healthcare (Li et al., 2023; Liu et al., [n. d.]; Holmes et al., 2023), industry (Dou et al., 2023), agriculture (Rezayi et al., 2023), law (Bubeck et al., 2023), gaming (Bubeck et al., 2023), and finance (Wu et al., 2023c), catalyzing a popular wave in AI (Liu et al., 2023a, g, h). • Rishi Bommasani, Drew A Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, et al. 2021.On the opportunities and risks of foundation models.arXiv preprint arXiv:2108.07258 (2021).
- 9. Sensory Inputs Sensory inputs refer to the various forms of data collected from different senses such as vision, hearing, touch, and smell that are processed by multimodal AI technology for a technical seminar. Data Fusion Data fusion involves combining information from multiple modalities, such as text, images, and videos, to improve the accuracy and robustness of AI systems in a technical seminar on multimodal AI technology generation. Machine Learning Algorithms Machine learning algorithms play a crucial role in generating multimodal AI technology for technical seminars by effectively analyzing and interpreting data from multiple sources such as text, images, and audio. Natural Language Processing Natural Language Processing is a crucial component of Multimodal AI technology, allowing for the analysis and understanding of human language in combination with other modalities such as images or videos. Computer Vision Computer Vision is a key component of Multimodal AI technology, which allows for the integration of visual data processing with other modes of information to enhance overall system performance.
- 12. Applications • Social media content moderation: Multimodal AI can be used to analyze text, images, and audio to identify and moderate harmful content on social media platforms. For instance, it can detect hate speech, violence, and bullying. • Virtual assistants: Smart assistants like Google Assistant and Amazon Alexa are powered by multimodal AI. They can understand and respond to natural language commands, both spoken and typed. • Healthcare imaging: In healthcare, multimodal AI can analyze medical images (X-rays, MRIs) along with text reports and patient history data to improve diagnostics. This can lead to more accurate diagnoses and better patient outcomes. • Autonomous vehicles: Self-driving cars rely heavily on multimodal AI. They use a variety of sensors, including cameras, radar, and LiDAR, to perceive their surroundings and navigate safely. • E-commerce product recommendations: Many e-commerce websites use multimodal AI to personalize product recommendations for customers. By considering both the product image and description, the AI can recommend items that are more likely to interest the customer
- 15. Conclusion • The future of AI is not just about seeing or hearing, it's about truly understanding. Multimodal AI holds the key to unlocking a new level of human-computer interaction, with applications that can bridge communication gaps, enhance our understanding of the world, and empower us to solve complex challenges in entirely new ways. The potential for positive impact across various fields is truly limitless.
- 16. References • Rania Abdelghani, Yen-Hsiang Wang, Xingdi Yuan, Tong Wang, Pauline Lucas, Hélène Sauzéon, and Pierre-Yves Oudeyer. 2023.GPT-3-driven pedagogical agents for training children’s curious question-asking skills. International Journal of Artificial Intelligence in Education 167, 3 (2023), 102887. • Hang Bao, Wen Wang, Li Dong, Qianru Liu, Ola K. Mohammed, Kirti Aggarwal, and Fang Wei. 2022.Vlmo: Unified vision-language pre-training with mixture-of-modality-experts. In Advances in Neural Information Processing Systems (NeurIPS), Vol. 35. 32897– 32912.