Dong Tian

Dong Tian

Auto-generated profile, based on publicly available information

Dong Tian

Mitsubishi Electric Research Laboratories (MERL)

Research Focus & Metrics

Academic Career Length

14Years (2013-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 November 2017 in IEEE Transactions on Signal Processing
Verified
Journal Article
Closed access
SCSiheng Chen
Siheng Chen
DTDong Tian
Dong Tian
CFChen Feng
Chen Feng
AVAnthony Vetro
Anthony Vetro
JKJelena Kovacevic
Jelena Kovacevic
To reduce cost in storing, processing and visualizing a large-scale point cloud, we consider a randomized resampling strategy to select a representative subset of points while preserving application-dependent features. The proposed strategy is based on graphs, which can represent underlying surfaces and lend themselves well to efficient computation. We use a general feature-extraction operator to represent application-dependent features and propose a general reconstruction error to evaluate the quality of resampling. We obtain a general form of optimal resampling distribution by minimizing the reconstruction error. The proposed optimal resampling distribution is guaranteed to be shift, rotation and scale-invariant in the 3D space. We next specify the feature-extraction operator to be a graph filter and study specific resampling strategies based on all-pass, low-pass, high-pass graph filtering and graph filter banks. We finally apply the proposed methods to three applications: large-scale visualization, accurate registration and robust shape modeling. The empirical performance validates the effectiveness and efficiency of the proposed resampling methods.

14 September 2021 in IEEE Transactions on Multimedia
Verified
Journal Article
Closed access
WHWei Hu
Wei Hu
JPJiahao Pang
Jiahao Pang
XLXianming Liu
Xianming Liu
DTDong Tian
Dong Tian
CLChia-Wen Lin
Chia-Wen Lin
AVAnthony Vetro
Anthony Vetro
Geometric data acquired from real-world scenes, e.g., 2D depth images, 3D point clouds, and 4D dynamic point clouds, have found a wide range of applications including immersive telepresence, autonomous driving, surveillance, etc. Due to irregular sampling patterns of most geometric data, traditional image/video processing methodologies are limited, while Graph Signal Processing (GSP)---a fast-developing field in the signal processing community---enables processing signals that reside on irregular domains and plays a critical role in numerous applications of geometric data from low-level processing to high-level analysis. To further advance the research in this field, we provide the first timely and comprehensive overview of GSP methodologies for geometric data in a unified manner by bridging the connections between geometric data and graphs, among the various geometric data modalities, and with spectral/nodal graph filtering techniques. We also discuss the recently developed Graph Neural Networks (GNNs) and interpret the operation of these networks from the perspective of GSP. We conclude with a brief discussion of open problems and challenges.

3 April 2018 in arXiv
Verified
Preprint
Open access
PDF file
YYYaoqing Yang
Yaoqing Yang
CFChen Feng
Chen Feng
YSYiru Shen
Yiru Shen
DTDong Tian
Dong Tian
Recent deep networks that directly handle points in a point set, e.g., PointNet, have been state-of-the-art for supervised learning tasks on point clouds such as classification and segmentation. In this work, a novel end-to-end deep auto-encoder is proposed to address unsupervised learning challenges on point clouds. On the encoder side, a graph-based enhancement is enforced to promote local structures on top of PointNet. Then, a novel folding-based decoder deforms a canonical 2D grid onto the underlying 3D object surface of a point cloud, achieving low reconstruction errors even for objects with delicate structures. The proposed decoder only uses about 7% parameters of a decoder with fully-connected neural networks, yet leads to a more discriminative representation that achieves higher linear SVM classification accuracy than the benchmark. In addition, the proposed decoder structure is shown, in theory, to be a generic architecture that is able to reconstruct an arbitrary point cloud from a 2D grid. Our code is available at http://www.merl.com/research/license#FoldingNet

8 September 2015 in arXiv
Verified
Preprint
Open access
PDF file
AGAkshay Gadde
Akshay Gadde
AKAndrew Knyazev
Andrew Knyazev
DTDong Tian
Dong Tian
HMHassan Mansour
Hassan Mansour
We study the problem of reconstructing a signal from its projection on a subspace. The proposed signal reconstruction algorithms utilize a guiding subspace that represents desired properties of reconstructed signals. We show that optimal reconstructed signals belong to a convex bounded set, called the "reconstruction" set. We also develop iterative algorithms, based on conjugate gradient methods, to approximate optimal reconstructions with low memory and computational costs. The effectiveness of the proposed approach is demonstrated for image magnification, where the reconstructed image quality is shown to exceed that of consistent and generalized reconstruction schemes.

26 September 2013
Verified
Conference Paper
Closed access
DTDong Tian
Dong Tian
FZFeng Zou
Feng Zou
CLChris Lee
Chris Lee
AVAnthony Vetro
Anthony Vetro
HSHuifang Sun
Huifang Sun
Depth-based 3D formats are currently being developed as extensions to both AVC and HEVC standards. The availability of depth information facilitates the generation of intermediate views for advanced 3D applications and displays, and also enables more efficient coding of the multiview input data through view synthesis prediction techniques. This paper outlines several approaches that have been explored to realize view synthesis prediction in modern video coding standards such as AVC and HEVC. The benefits and drawbacks of various architectures are analyzed in terms of performance, complexity, and other design considerations. It is hence concluded that block-based VSP prediction for multiview video signals provides attractive coding gains with comparable complexity as traditional motion/disparity compensation.

5 September 2015 in arXiv
Verified
Preprint
Open access
PDF file
DTDong Tian
Dong Tian
HMHassan Mansour
Hassan Mansour
AKAndrew Knyazev
Andrew Knyazev
AVAnthony Vetro
Anthony Vetro
In 3D image/video acquisition, different views are often captured with varying noise levels across the views. In this paper, we propose a graph-based image enhancement technique that uses a higher quality view to enhance a degraded view. A depth map is utilized as auxiliary information to match the perspectives of the two views. Our method performs graph-based filtering of the noisy image by directly computing a projection of the image to be filtered onto a lower dimensional Krylov subspace of the graph Laplacian. We discuss two graph spectral denoising methods: first using Chebyshev polynomials, and second using iterations of the conjugate gradient algorithm. Our framework generalizes previously known polynomial graph filters, and we demonstrate through numerical simulations that our proposed technique produces subjectively cleaner images with about 1-3 dB improvement in PSNR over existing polynomial graph filters.