Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore how generative AI and visual language models are revolutionizing printers and scanners into intelligent document assistants in this 29-minute conference talk. Learn about the groundbreaking deployment of multimodal AI models like QN 2.5 VL and LayoutLMv3 directly on edge devices, enabling powerful capabilities such as document summarization, translation, layout adaptation, and ad removal—all processed on-device without requiring cloud connectivity. Discover how combining 4-bit quantization, image downscaling, and NVIDIA's optimized libraries allows even memory-constrained devices with 16GB GPUs to process and enhance documents in real-time. Understand how this approach not only boosts performance and privacy but fundamentally reshapes how enterprises handle sensitive data by transforming passive printers into intelligent document processing systems operating at the edge.
