|
My name is Trong-Thang Pham. I got my bachelor's degree in Computer Science at University of Science, VNU-HCM, Vietnam. Currently, I'm working as AI Researcher at AIOZ AI and studying a master's degree in Artificial Intelligence at John von Neumann Institute, VNU-HCM. At AIOZ AI and my master's study, I've worked mainly on photorealistic 2D and 3D reconstructions. My avatar on the left was enhanced using GAN. For 3D reconstruction, I am working on both template-free problems (e.g., recontructing a bird from a monocular RGB video with its masks) and human-specific problems (e.g., reconstructing a sequence of 3D human models from a monocular RGB video). |
|
|
I'm interested in 3D reconstruction, representation learning, probabilistic deep learning, deep learning theory, and their applications in real life. |
|
Minh-Triet Tran, Tam V Nguyen, Trung-Hieu Hoang, Trung-Nghia Le, Khac-Tuan Nguyen, Dat-Thanh Dinh, Thanh-An Nguyen, Hai-Dang Nguyen, Xuan-Nhat Hoang, Trong-Tung Nguyen, Viet-Khoa Vo-Ho, Trong-Le Do, Lam Nguyen, Minh-Quan Le, Hoang-Phuc Nguyen-Dinh, Trong-Thang Pham, Xuan-Vy Nguyen, E-Ro Nguyen, Quoc-Cuong Tran, Hung Tran, Hieu Dao, Mai-Khiem Tran, Quang-Thuc Nguyen, Tien-Phat Nguyen, Gia-Han Diep, Minh N Do CVPRW , 2020
We introduce our Intelligent Traffic Analysis Software Kit (iTASK) to tackle three challenging problems: vehicle flow counting, vehicle re-identification, and abnormal event detection. |
|
Minh-Khoi Pham, Hai-Tuan Ho-Nguyen, Trong-Thang Pham, Hung Vinh Tran, Hai-Dang Nguyen, Minh-Triet Tran Pixel Privacy track at Multimedia Eval Workshop , 2020
In this challenge, we have implemented methods of combining image enhancement and an end to end attack. The final results show that all of our approaches successfully fool the BIQA system.
|
|
Hung Vinh Tran*, Trong-Thang Pham* , Hai-Tuan Ho-Nguyen*, Hoai-Lam Nguyen-Hy*, Xuan-Vy Nguyen*, Thang-Long Nguyen-Ho*, Minh-Triet Tran Pixel Privacy track at Multimedia Eval Workshop , 2019
We propose the method to attack the ResNet50 model using back propagation. For each input image, we consider the logit for its target class as a function on the input image's pixels. This allows back propagation to minimize this logit by modifying the input image. We also use several methods to enhance the input images' quality before feeding it through the protection phase.
|
|
Website template from Jon Barron. |