Industrial Technology Research Institute

Ai.skopy, Inc.

Chih-Chung Hsu

Chia-Wen Lin

Tsung-Chun Lee

A system and method of diagnosing gastrointestinal neoplasm or pathologies in an endoscopy system including an endoscopy system display for displaying an image enhanced endoscopy (IEE) image. The method includes randomly generating training image samples with or without cancer region(s) by an adversarial network (AN) including collecting endoscopic training images (T1) and automatically generating a realistic IEE image as a new training image sample (T2) using a generator network in the AN from a generated segmentation map; using a prediction network (L1PN) to learn a level 1 prediction result being a cancerous probability of an IEE image from the collected T1 and T2; using a prediction network (L2PN) to learn a level 2 prediction result being detected cancerous region(s) of an IEE image; and predicting the level 1 result and the level 2 result for an IEE image using the L1PN and the L2PN and without using the AN.

System and method for diagnosing gastrointestinal neoplasm

A method for example-based face hallucination uses manifold learning to project a plurality of training images in a training database and an input low resolution (LR) face image into a same manifold domain, then iteratively refines the reconstruction basis by selecting a training set having k projected training images which best match the parts of the projected LR face image, where k&#x2266;N and N is the number of projected training images. Through the best-match training set, a set of prototype faces are learned, and the set of prototype faces are used as the reconstruction basis to reconstruct a high resolution face image for the input LR face image.

Growing community of inventors

Taichung, Taiwan

Chih-Chung Hsu