Ching-Wei Tseng Email & Phone Number

Hsinchu, TW

• Develop innovative algorithm for AR/VR devices
• Develop and Release 3D segmentation solution for target XR devices.
• Transfer learning from foundation/large multi-modal model to on-device model.
• Leverage 3D priors in networks and handle different camera orientations.
• Global collaborations: data acquisition, self-supervised depth estimation, uncertainty, end-to-end 3D reconstruction

台灣 Taiwan 新竹市

• Developing algorithm for mobile camera features
• Deep Learning Model Development: - Stereo Depth Estimation, Human Matting/Segmentation, Depth Completion/Refinement
• Model Optimization: - Handcrafted Pruning, Quantization Aware Training (QAT), Model Porting to Device

台灣 Taiwan 新竹市

Taiwan

• Proposed a general image completion method by using deep auto-encoder structures and adversarial training (GANs) techniques.
• Proposed a depth super-resolution algorithm via multi-frame fusion and guided image filtering with end-to-end deep learning.
• Developing a mental workload assessment system based on human action recognition with Kinect1.0/2.0

Taiwan

Master Of Science (M.S.), Institute Of Computer Science And Engineering

● Computer Vision Lab, led by Prof. Shang-Hong Lai ● TA in Introduction to Multimedia (10420CS 357000)

Bachelor Of Science (B.S.), Department Of Industrial Engineering And Engineering Management, 3.83

Activities and Societies: IEEM Summer Camp and Student Association● Vision & Learning Lab, led by Prof. Szu-Hao Huang. ● Minor in Computer.