Abstract: Humans can intuit answers to visual problems instantly without too much thinking. But it has always been challenging to develop intuition in deep learning modeling to solve more complex problems like OCR and document understanding. So how can you transfer the human learning process to deep learning models? The OCI Vision team has developed a technology process that has been significantly more effective in addressing practical problems such as model selection and development in object detection, OCR, and document understanding. Attend this talk to:
· Explore different types of intuition problems in computer vision
· Discover multiple approaches to solving this problem
· Learn about Oracle’s solution to intuition in deep learning
Bio: As Senior Director of Machine Learning at Oracle, Jun Qian leads a team building deep learning vision services across the areas of image, document, and video. He has over 20 years of experience in the AI world and has been instrumental in creating ML and AI solutions at Amazon and Microsoft. Outside of work, Jun lives in the Seattle area and likes playing with and learning from his two curious kids.