Shen Zheng

I am currently a Perception Software Engineer working at Lucid Motors . Recently, I completed my Master of Science in Computer Vision (MSCV) degree at Carnegie Mellon University (CMU), collaborating with Dr. Srinivasa Narasimhan at the Illumination and Imaging Lab (LLIM). Prior to joining CMU, I earned my bachelor's degree in Mathematics from Wenzhou-Kean University (WKU), where I worked with Dr. Gaurav Gupta .

Email: shenzhen@andrew.cmu.edu

CV  /  Google Scholar  /  Github  /  Linkedin  /  Leetcode  /  YouTube

My research focus is better scene understanding across diverse environments using the following strategies:

• Test-Time Preprocessing: Blind Motion Deblurring, Single Image Deraining, Low-Light Image Enhancement
• Train-Time Preprocessing: Image Warping
• Finetuning: Image Generation, Image-to-Image Translation

Motivation: Object Scale bias challenges contemporal visual recognition systems.

Solution: Propose a instance-level image warping technique using dataset-specific size statistics to warp images in-place during training to address object scale bias. Integrate image warping and feature unwarping into domain adaptation in a task-agnostic way without warping at test time.

Motivation: Previous image-to-image translation methods produce artifacts and distortions, and lack control over the amount of rain generated.

Solution: Introduce a Triangular Probability Similarity (TPS) loss to minimize the artifacts and distortions during rain generation. Propose a Semantic Noise Contrastive Estimation (SeNCE) strategy to optimize the amounts of generated rain. Evaluate rain generation performances using rain removal and object detection.

Motivation: Existing LLIE datasets focus on either overexposure or underexposure, not both, and usually feature minimally degraded images captured from static positions.

Solution: Present a comprehensive survey of low-light image enhancement (LLIE). Propose the SICE_Grad and SICE_Mix image datasets, which include images with both overexposure and underexposure. Introduce Night Wenzhou, a large-scale, high-resolution video dataset captured in fast motion with diverse illuminations and degradation.

Motivation: Current point cloud analysis methods struggles with irregular (i.e., unevenly distributed) point clouds.

Solution: PointNorm, a point cloud analysis network with a DualNorm module (Point Normalization & Reverse Point Normalization) that leverages local mean and global standard deviation.

Motivation: Current low-light image enhancement methods cannot handle uneven illuminations, is computationally inefficient, and fail to preserve the semantic information.

Solution: SGZ, a zero-shot low-light image enhancement framework with pixel-wise light deficiency estimation, parameter-free recurrent image enhancement, and unsupervised semantic segmentation.

Motivation: Generative models experience posterior collapse and vanishing gradient due to no effective metric for real-fake image evaluation.

Solution: Propose Adversarial Similarity Distance Introspective Variational Autoencoder (AS-IntroVAE), which can address the posterior collapse and the vanishing gradient problem in image generation in one go.

Motivation: Object detection algorithms exhibit suboptimal performance on blurry scenes.

Solution: Propose Deblur-YOLO, a generative adversarial network with a dilated feature pyramid generator, double multi-scale discriminators, and a detection discriminator to deal with photographs corrupted by motion blur.

Motivation: Existing ConvNets have poor computational complexity and require significant memory consumption.

Solution: Introduce an efficient ConvNet with weighted average stacking, Winograd-ReLU-based network pruning, and a electromagnetic-inspired dynamic batch size algorithm.

Motivation: Former deraining approaches often eliminate essential background details along with the rain, hindering tasks such as detection and segmentation.

Solution: SAPNet, an image-deraining network that integrates low-level image-deraining and high-level background segmentation using progressive dilated unit, perceptual contrastive loss, and unsupervised background segmentation.

Motivation: Previous unsupervised domain adaptation methods for medical imaging falter across varied imaging modalities due to substantial domain differences.

Solution: UDA-VAE++, an unsupervised domain adaptation framework that leverages mutual information maximization and sequential reparameterization for cardiac segmentation.

Director: Dr. Wangjiang Zhu

Responsible for long-tailed data augmentation, training data auto-labeling and cleaning, and model evaluation for traffic light detection algorithms.

Implemented CycleGAN to conduct unsupervised data augmentation, converting traffic light bulbs from left arrow to round & leftUturn arrow.

Constructed a traffic light auto-label model using quantized VoVNet-57, filtering 14,618 incorrect annotations from 1,160,513 labeled frames.

Increased the classification accuracy for leftUturn traffic light from 78.41% to 87.27%, and the mean average precision from 93.01% to 94.80%.