181 citations · 320 across the 15 of their papers we have counts for
14 papers · 1 filter
Mitigating Task Interference in Multi-Task Learning via Explicit Task Routing with Non-Learnable Primitives
Chuntao Ding, Zhichao Lu, Shangguang Wang +2
Multi-task learning (MTL) seeks to learn a single model to accomplish multiple tasks by leveraging shared information among the tasks. Existing MTL models, however, have been known…
Revisiting Residual Networks for Adversarial Robustness: An Architectural Perspective
Shihua Huang, Zhichao Lu, Kalyanmoy Deb +1
Efforts to improve the adversarial robustness of convolutional neural networks have primarily focused on developing more effective adversarial training methods. In contrast, little…
Surrogate-assisted Multi-objective Neural Architecture Search for Real-time Semantic Segmentation
Zhichao Lu, Ran Cheng, Shihua Huang +3
The architectural advancements in deep neural networks have led to remarkable leap-forwards across a broad array of computer vision tasks. Instead of relying on human expertise, ne…
Multiview Transformers for Video Recognition
Shen Yan, Xuehan Xiong, Anurag Arnab +4
Video understanding requires reasoning at multiple spatiotemporal resolutions -- from short fine-grained motions to events taking place over longer durations. Although transformer…
FaPN: Feature-aligned Pyramid Network for Dense Image Prediction
Shihua Huang, Zhichao Lu, Ran Cheng +1
Recent advancements in deep neural networks have made remarkable leap-forwards in dense image prediction. However, the issue of feature alignment remains as neglected by most exist…
End-to-End Dense Video Captioning with Parallel Decoding
Teng Wang, Ruimao Zhang, Zhichao Lu +3
Dense video captioning aims to generate multiple associated captions with their temporal locations from the video. Previous methods follow a sophisticated "localize-then-describe"…
The surprising impact of mask-head architecture on novel class segmentation
Vighnesh Birodkar, Zhichao Lu, Siyang Li +2
Instance segmentation models today are very accurate when trained on large annotated datasets, but collecting mask annotations at scale is prohibitively expensive. We address the p…
Multi-objective Neural Architecture Search with Almost No Training
Shengran Hu, Ran Cheng, Cheng He +1
In the recent past, neural architecture search (NAS) has attracted increasing attention from both academia and industries. Despite the steady stream of impressive empirical results…
NSGANetV2: Evolutionary Multi-Objective Surrogate-Assisted Neural Architecture Search
Zhichao Lu, Kalyanmoy Deb, Erik Goodman +2
In this paper, we propose an efficient NAS algorithm for generating task-specific models that are competitive under multiple competing objectives. It comprises of two surrogates, o…
Learning Video Representations from Textual Web Supervision
Jonathan C. Stroud, Zhichao Lu, Chen Sun +4
Videos on the Internet are paired with pieces of text, such as titles and descriptions. This text typically describes the most important content in the video, such as the objects i…
Neural Architecture Transfer
Zhichao Lu, Gautam Sreekumar, Erik Goodman +3
Neural architecture search (NAS) has emerged as a promising avenue for automatically designing task-specific neural networks. Existing NAS approaches require one complete search fo…
MUXConv: Information Multiplexing in Convolutional Neural Networks
Zhichao Lu, Kalyanmoy Deb, Vishnu Naresh Boddeti
Convolutional neural networks have witnessed remarkable improvements in computational efficiency in recent years. A key driving force has been the idea of trading-off model express…
Multi-Objective Evolutionary Design of Deep Convolutional Neural Networks for Image Classification
Zhichao Lu, Ian Whalen, Yashesh Dhebar +4
Early advancements in convolutional neural networks (CNNs) architectures are primarily driven by human expertise and by elaborate design processes. Recently, neural architecture se…
NSGA-Net: Neural Architecture Search using Multi-Objective Genetic Algorithm
Zhichao Lu, Ian Whalen, Vishnu Boddeti +4
This paper introduces NSGA-Net -- an evolutionary approach for neural architecture search (NAS). NSGA-Net is designed with three goals in mind: (1) a procedure considering multiple…