|
I am a computer science PhD candidate at University of California, Los Angeles (UCLA), and a researcher at David Geffen School of Medicine and UCLA Stein Eye Institute . In my PhD journey, I am honored to be advised by Demetri Terzopoulos. My research interests span across computer vision, deep learning and artificial intelligence. I am the recipient of the 2018 UCLA Henry Samueli School of Engineering and Applied Science (HSSEAS) Edward K. Rice Outstanding Masters Student Award. Email / Google Scholar / LinkedIn / GitHub |
|
Collaborators
Awards
|
ResearchMy research focus lies at the intersection of computer vision, machine learning, and Artificial Intelligence (AI) and combines computational vision with machine intelligence to create a new class of robust representation learning models that are more resilient to outliers and require less training data. These systems hold tremendous promise for applications such as healthcare (e.g. medical image analysis), autonomous vehicles and remote-sensing, where brittle, black-box AI models are failing to find traction. |
 |
Multimodal brain tumor segmentation challenge (BraTS) brings together researchers to improve automated methods for 3D MRI brain tumor segmentation. Tumor segmentation is one of the fundamental vision tasks necessary for diagnosis and treatment planning of the disease. Previous years winning methods were all deep-learning based, thanks to the advent of modern GPUs, which allow fast optimization of deep convolutional neural network architectures. In this work, we explore best practices of 3D semantic segmentation, including conventional encoder-decoder architecture, as well combined loss functions, in attempt to further improve the segmentation accuracy. We evaluate the method on BraTS 2019 challenge. |
|||||||||
 |
The Active Contour Model (ACM) is a standard image analysis technique whose numerous variants have attracted an enormous amount of research attention across multiple fields. Incorrectly, however, the ACM's differential-equation-based formulation and prototypical dependence on user initialization have been regarded as being largely incompatible with the recently popular deep learning approaches to image segmentation. This paper introduces the first tight unification of these two paradigms. In particular, we devise Deep Convolutional Active Contours (DCAC), a truly end-to-end trainable image segmentation framework comprising a Convolutional Neural Network (CNN) and an ACM with learnable parameters. The ACM's Eulerian energy functional includes per-pixel parameter maps predicted by the backbone CNN, which also initializes the ACM. Importantly, both the CNN and ACM components are fully implemented in TensorFlow, and the entire DCAC architecture is end-to-end automatically differentiable and backpropagation trainable without user intervention. As a challenging test case, we tackle the problem of building instance segmentation in aerial images and evaluate DCAC on two publicly available datasets, Vaihingen and Bing Huts. Our reseults demonstrate that, for building segmentation, the DCAC establishes a new state-of-the-art performance by a wide margin. |
|||||||||
 |
Automated segmentation of kidneys and kidney tumors is an important step in quantifying the tumor's morphometrical details to monitor the progression of the disease and accurately compare decisions regarding the kidney tumor treatment. Manual delineation techniques are often tedious, error-prone and require expert knowledge for creating unambiguous representation of kidneys and kidney tumors segmentation. In this work, we propose an end-to-end boundary aware fully Convolutional Neural Networks (CNNs) for reliable kidney and kidney tumor semantic segmentation from arterial phase abdominal 3D CT scans. We propose a segmentation network consisting of an encoder-decoder architecture that specifically accounts for organ and tumor edge information by devising a dedicated boundary branch supervised by edge-aware loss terms. We have evaluated our model on 2019 MICCAI KiTS Kidney Tumor Segmentation Challenge dataset and our method has achieved dice scores of 0.9742 and 0.8103 for kidney and tumor repetitively and an overall composite dice score of 0.8923. |
|||||||||
 |
Fully convolutional neural networks (CNNs) have proven to be effective at representing and classifying textural information, thus transforming image intensity into output class masks that achieve semantic image segmentation. In medical image analysis, however, expert manual segmentation often relies on the boundaries of anatomical structures of interest. We propose boundary aware CNNs for medical image segmentation. Our networks are designed to account for organ boundary information, both by providing a special network edge branch and edge-aware loss terms, and they are trainable end-to-end. We validate their effectiveness on the task of brain tumor segmentation using the BraTS 2018 dataset. Our experiments reveal that our approach yields more accurate segmentation results, which makes it promising for more extensive application to medical image segmentation. |
|||||||||
 |
Lesion segmentation is an important problem in computer-assisted diagnosis that remains challenging due to the prevalence of low contrast, irregular boundaries that are unamenable to shape priors. We introduce Deep Active Lesion Segmentation (DALS), a fully automated segmentation framework for that leverages the powerful nonlinear feature extraction abilities of fully Convolutional Neural Networks (CNNs) and the precise boundary delineation abilities of Active Contour Models (ACMs). Our DALS framework benefits from an improved level-set ACM formulation with a per-pixel-parameterized energy functional and a novel multiscale encoder-decoder CNN that learns an initialization probability map along with parameter maps for the ACM. We evaluate our lesion segmentation model on a new Multiorgan Lesion Segmentation (MLS) dataset that contains images of various organs, including brain, liver, and lung, across different imaging modalities---MR and CT. Our results demonstrate favorable performance compared to competing methods, especially for small training datasets. |
|||||||||
 |
The reliable segmentation of retinal vasculature can provide the means to diagnose and monitor the progression of a variety of diseases affecting the blood vessel network, including diabetes and hypertension. We leverage the power of convolutional neural networks to devise a reliable and fully automated method that can accurately detect, segment, and analyze retinal vessels. In particular, we propose a novel, fully convolutional deep neural network with an encoder-decoder architecture that employs dilated spatial pyramid pooling with multiple dilation rates to recover the lost content in the encoder and add multiscale contextual information to the decoder. We also propose a simple yet effective way of quantifying and tracking the widths of retinal vessels through direct use of the segmentation predictions. Unlike previous deep-learning-based approaches to retinal vessel segmentation that mainly rely on patch-wise analysis, our proposed method leverages a whole-image approach during training and inference, resulting in more efficient training and faster inference through the access of global content in the image. We have tested our method on two publicly available datasets, and our state-of-the-art results on both the DRIVE and CHASE-DB1 datasets attest to the effectiveness of our approach. |
|||||||||
 |
This work presents the first end-to-end 3D deep learning-based lung lobe segmentation model which sets a new state-of-the-art record.The proposed model has accomplished overall Dice scores of 0.94 and 0.95 on LIDC and LTRC data-sets while running under 2 seconds on a single NVIDIA Titan XP GPU during inference. This paper has received the NVIDIA Best Paper Award at the MICCAI Workshop on Deep Learning for Medical Image Analysis . |
|||||||||
 |
This paper builds upon the concept of Boundary Learning Optimization Tool (BLOT) and introduces a new approach for automatic hyper-parameter initialization as well as a novel optimization algorithm for tracing the performance boundaries by leveraging Sequential Quadratic Programming(SQP) and Augmented Lagrangian Pattern Search(ALPS). In comparison to the original BLOT paper, it is more efficient and equipped to rigorously search conflicting boundary regions. |
|||||||||
 |
In this work, the first deep learning-based CAD framework called the Boundary Learning Optimization Tool (BLOT) has been introduced. BLOT helps designers to rapidly explore the design space of their synthesized topologies to identify the corresponding optimal design instantiations by creating physical simulations and leveraging the power of non-linear optimization and deep learning. This paper has received the Theoretical Contributions in Compliant Mechanisms Award at (ASME) Mechanisms and Robotics Conference in the International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE). |
Teaching
|
