Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn how synthetic data generation can revolutionize industrial computer vision deployment from a Meta Applied Research Scientist in this 18-minute conference talk. Discover why 54% of AI projects fail at the proof-of-concept stage due to data acquisition challenges, particularly in manufacturing and industrial automation where collecting images for object detection can take 6-12 months. Explore how generative AI transforms this bottleneck by enabling synthetic data creation that accelerates development cycles and enables faster deployment of robust AI models in production environments. Gain insights from an expert who works with billion-scale datasets to train massive multimodal foundational models at Facebook AI Research, with previous experience at Citadel as a Quantitative Researcher and Amazon Alexa AI on large-scale multimodal models and Embodied AI applications. Understand the practical applications of self-supervised learning and multimodal AI in solving real-world industrial computer vision challenges.
