Academic Profile : Faculty
Prof Lin Weisi
Associate Dean (Research), College of Computing & Data Science
President's Chair in Computer Science
Professor, College of Computing & Data Science
Email
Weisi Lin graduated from Sun Yat-Sen (Zhongshan) University, China with B.Sc in Electronics and M.Sc in Digital Signal Processing, respectively, and from King’s College, London University, UK with Ph.D in Computer Vision. He taught and researched in Sun Yat-Sen University, Shantou University (China), Bath University (UK), National University of Singapore, Institute of Microelectronics (Singapore), and Institute for Infocomm Research (Singapore). He has been the project leader of 10+ successfully delivered projects in digital multimedia technology development. He also served as the Lab Head, Visual Processing, and then the Acting Department Manager, Media Processing, in Institute for Infocomm Research. He holds 16 patents, wrote 9 book chapters, edited 3 books, authored 2 books, published over 400 refereed papers in international journals and conferences, and made more than ten contributions to international standardization. He believes that good theory is practical so has kept a balance of academic research and industrial deployment throughout his working life.
perception-inspired signal modeling, visual quality evaluation, video compression, image processing & analysis, multimedia systems
- Climate Transformation Programme - Cross-cutting theme 3 (Weisi Lin)
- Exploring visual signal representation towards machine uses
- Language-Guided Controllable Aesthetic Image Editing
- Overheads M4 Account
- President's Chair in Computer Science
- Revisiting UGC Dataset and Benchmarking Metric
- Smart Technologies Lab
- WP4: Scent Digitalization and Computation (SDC)- PI: Lin Weisi
US 2014/0301462 A1: Lossless Image and Video Compression (2017)
Abstract: A method is provided for encoding an intra predicted residual block of an image for use in image or video compression. The intra predicted residual block is associated with an intra prediction coding mode. The method includes generating a set of residual error blocks including residual data with different statistical characteristics from the residual data in the intra predicted residual block. Each of the residual error blocks is scanned and entropy coded to produce a first set of bit streams. The lengths of each of the first set of bit streams are recorded. The intra predicted residual block is also scanned and entropy coded to produce a second bit stream. The length of the second bit stream is recorded. Selecting the minimum length bit stream from the first set of bit streams and the second bit stream as the output coded bit stream of the intra predicted residual block.
US 2014/0169662 A1: Image Retargeting Quality Assessment (2015)
Abstract: A method of performing an image retargeting quality assessment comprising comparing an original image and a retargeted image in a frequency domain, wherein the retargeted image is obtained by performing a retargeting algorithm on the original image. The disclosure also includes an apparatus comprising a processor configured to perform an image retargeting quality assessment, and compare an original image and a retargeted image in a spatial domain, wherein the retargeted image is obtained by performing a retargeting algorithm on the original image, and wherein comparing the original image and the retargeted image in the spatial domain comprises comparing the original image and the retargeted image to determine an amount of shape distortion between the images..
US 2023/0075888 A1: Object Re-Identification Using Multiple Cameras (2024)
Abstract: In some aspects, a method for object re-identification may include obtaining a first set of images from a first camera, and a second set of images from at least one second camera; determining a first set of features based on the first set of images, the first set of features lying in a first feature space; and determining a second set of features based on the second set of images, the second set of features lying in a second feature space. The method may additionally include determining a first feature projection matrix and a second feature projection matrix that respectively map the first set of features and the second set of features to a shared feature space; and determining a common dictionary based on the shared feature space.
Abstract: A method is provided for encoding an intra predicted residual block of an image for use in image or video compression. The intra predicted residual block is associated with an intra prediction coding mode. The method includes generating a set of residual error blocks including residual data with different statistical characteristics from the residual data in the intra predicted residual block. Each of the residual error blocks is scanned and entropy coded to produce a first set of bit streams. The lengths of each of the first set of bit streams are recorded. The intra predicted residual block is also scanned and entropy coded to produce a second bit stream. The length of the second bit stream is recorded. Selecting the minimum length bit stream from the first set of bit streams and the second bit stream as the output coded bit stream of the intra predicted residual block.
US 2014/0169662 A1: Image Retargeting Quality Assessment (2015)
Abstract: A method of performing an image retargeting quality assessment comprising comparing an original image and a retargeted image in a frequency domain, wherein the retargeted image is obtained by performing a retargeting algorithm on the original image. The disclosure also includes an apparatus comprising a processor configured to perform an image retargeting quality assessment, and compare an original image and a retargeted image in a spatial domain, wherein the retargeted image is obtained by performing a retargeting algorithm on the original image, and wherein comparing the original image and the retargeted image in the spatial domain comprises comparing the original image and the retargeted image to determine an amount of shape distortion between the images..
US 2023/0075888 A1: Object Re-Identification Using Multiple Cameras (2024)
Abstract: In some aspects, a method for object re-identification may include obtaining a first set of images from a first camera, and a second set of images from at least one second camera; determining a first set of features based on the first set of images, the first set of features lying in a first feature space; and determining a second set of features based on the second set of images, the second set of features lying in a second feature space. The method may additionally include determining a first feature projection matrix and a second feature projection matrix that respectively map the first set of features and the second set of features to a shared feature space; and determining a common dictionary based on the shared feature space.
Awards
Highly Cited Researcher 2019, 2020, 2021 (awarded by Clarivate Analytics)
Fellowships & Other Recognition
Fellow of IEEE, Fellow of IET, Honorary Fellow of Singapore Institute of Engineering Technologists, Chartered Engineer