I'm a PhD student in The Australian National University, with my interest focus on Computer Vision and Deep Learning, supervised by Hongdong Li, Antonio Robles-Kelly, Shaodi You and Yasuyuki Matsushita.
Previously, I received the Master degree in Australian National University, major in master of computing, in 2018. And B. Eng degree in Shanghai Jiaotong University in 2016, under the supervision of Decheng Wang.
Here's my CV.
Learning to Minify Photometric Stereo
Junxuan Li, Antonio Robles-Kelly, Shaodi You, and Yasuyuki Matsushita. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. (CVPR) 2019. [pdf] [code]
Photometric stereo estimates the surface normal given a set of images acquired under different illumination conditions. To deal with diverse factors involved in the image formation process, recent photometric stereo methods demand a large number of images as input. We propose a method that can dramatically decrease the demands on the number of images by learning the most informative ones under different illumination conditions. To this end, we use a deep learning framework to automatically learn the critical illumination conditions required at input. Furthermore, we present an occlusion layer that can synthesize cast shadows, which effectively improves the estimation accuracy. We assess our method on challenging real-world conditions, where we outperform techniques elsewhere in the literature with a significantly reduced number of light conditions.
A Frequency Domain Neural Network for Fast Image Super-resolution
Junxuan Li, Shaodi You, and Antonio Robles-Kelly. Neural Networks (IJCNN), 2018 International Joint Conference on. IEEE, 2018. Oral Presentation. [arXiv] [pdf] [code]
In this paper, we present a frequency domain neural network for image super-resolution. The network employs the convolution theorem so as to cast convolutions in the spatial domain as products in the frequency domain. Moreover, the non-linearity in deep nets, often achieved by a rectifier unit, is here cast as a convolution in the frequency domain. This not only yields a network which is very computationally efficient at testing but also one whose parameters can all be learnt accordingly. The network can be trained using back propagation and is devoid of complex numbers due to the use of the Hartley transform as an alternative to the Fourier transform. Moreover, the network is potentially applicable to other problems elsewhere in computer vision and image processing which are often cast in the frequency domain. We show results on super-resolution and compare against alternatives elsewhere in the literature. In our experiments, our network is one to two orders of magnitude faster than the alternatives with an imperceptible loss of performance.
Stereo Super-resolution via a Deep Convolutional Network
Junxuan Li, Shaodi You, and Antonio Robles-Kelly. Digital Image Computing: Techniques and Applications (DICTA), 2017 International Conference on. IEEE, 2017. Oral Presentation. [pdf]
In this paper, we present a method for stereo super-resolution which employs a deep network. The network is trained using the residual image so as to obtain a high resolution image from two, low resolution views. Our network is comprised by two deep sub-nets which share, at their output, a single convolutional layer. This last layer in the network delivers an estimate of the residual image which is then used, in combination with the left input frame of the stereo pair, to compute the super-resolved image at output. Each of these sub-networks is comprised by ten weight layers and, hence, allows our network to combine structural information in the image across image regions efficiently. Moreover, by learning the residual image, the network copes better with vanishing gradients and its devoid of gradient clipping operations. We illustrate the utility of our network for image-pair super-resolution and compare our network to its non-gradient trained analogue and alternatives elsewhere in the literature.
This is a project required by course ENGN6528 Computer Vision. FlowNet and several works on optical estimation using CNNs shows the potential ability of CNNs in doing per-pixel regression. We proposed several CNNs network architectures that can estimate optical flow, and fully unveiled the intrinsic different between these structures.
The interface development and application of OPT-Ship
Junxuan Li, Supervisor: Prof. Decheng Wang, 2016
This is my undergraduate graduation project thesis. It implemented the interface of a software - OPTShip - by using C++ and Qt platform.
Research of mass transit passenger flow distribution base on IC and GPS data
Junxuan Li, Supervisor: Dr. Linjie Gao, 2011
This is the 26th Participation in Research Program(PRP). This project was aimed to analysis the data retrieved from IC and GPS and give an overall judgment to transit distribution. It was completed by using Python.
modified from © Yihui He 2017