Searcharxiv⌕ Search

arXiv subjects

Shen Yan

Publications and source records attributed to Shen Yan.

80 records · Page 5Linked to original sources
arXiv2020

MutualNet: Adaptive ConvNet via Mutual Learning from Network Width and Resolution

We propose the width-resolution mutual learning method (MutualNet) to train a network that is executable at dynamic resource constraints to achieve adaptive accuracy-efficiency trade-offs at runtime. Our method trains a cohort of sub-networks with different widths using different input resolutions to mutually learn multi-scale representations for each sub-network. It achieves consistently better ImageNet top-1 accuracy over the state-of-the-art adaptive network US-Net under different computation constraints, and outperforms the best compound scaled MobileNet in EfficientNet by 1.5%. The superiority of our method is also validated on COCO object detection and instance segmentation as well as transfer learning. Surprisingly, the training strategy of MutualNet can also boost the performance of a single network, which substantially outperforms the powerful AutoAugmentation in both efficiency (GPU search hours: 15000 vs. 0) and accuracy (ImageNet: 77.6% vs. 78.6%). Code is available at \url{https://github.com/taoyang1122/MutualNet}.

cs.CV↗
arXiv2020

Bregman Augmented Lagrangian and Its Acceleration

We study the Bregman Augmented Lagrangian method (BALM) for solving convex problems with linear constraints. For classical Augmented Lagrangian method, the convergence rate and its relation with the proximal point method is well-understood. However, the convergence rate for BALM has not yet been thoroughly studied in the literature. In this paper, we analyze the convergence rates of BALM in terms of the primal objective as well as the feasibility violation. We also develop, for the first time, an accelerated Bregman proximal point method, that improves the convergence rate from $O(1/\sum_{k=0}^{T-1}η_k)$ to $O(1/(\sum_{k=0}^{T-1}\sqrt{η_k})^2)$, where $\{η_k\}_{k=0}^{T-1}$ is the sequence of proximal parameters. When applied to the dual of linearly constrained convex programs, this leads to the construction of an accelerated BALM, that achieves the improved rates for both primal and dual convergences.

math.OC↗
arXiv2020

Improve Unsupervised Domain Adaptation with Mixup Training

Unsupervised domain adaptation studies the problem of utilizing a relevant source domain with abundant labels to build predictive modeling for an unannotated target domain. Recent work observe that the popular adversarial approach of learning domain-invariant features is insufficient to achieve desirable target domain performance and thus introduce additional training constraints, e.g. cluster assumption. However, these approaches impose the constraints on source and target domains individually, ignoring the important interplay between them. In this work, we propose to enforce training constraints across domains using mixup formulation to directly address the generalization performance for target data. In order to tackle potentially huge domain discrepancy, we further propose a feature-level consistency regularizer to facilitate the inter-domain constraint. When adding intra-domain mixup and domain adversarial learning, our general framework significantly improves state-of-the-art performance on several important tasks from both image classification and human activity recognition.

stat.ML↗
arXiv2019

Unsupervised Visual Representation Learning with Increasing Object Shape Bias

(Very early draft)Traditional supervised learning keeps pushing convolution neural network(CNN) achieving state-of-art performance. However, lack of large-scale annotation data is always a big problem due to the high cost of it, even ImageNet dataset is over-fitted by complex models now. The success of unsupervised learning method represented by the Bert model in natural language processing(NLP) field shows its great potential. And it makes that unlimited training samples becomes possible and the great universal generalization ability changes NLP research direction directly. In this article, we purpose a novel unsupervised learning method based on contrastive predictive coding. Under that, we are able to train model with any non-annotation images and improve model's performance to reach state-of-art performance at the same level of model complexity. Beside that, since the number of training images could be unlimited amplification, an universal large-scale pre-trained computer vision model is possible in the future.

cs.CV↗
arXiv2019

HM-NAS: Efficient Neural Architecture Search via Hierarchical Masking

The use of automatic methods, often referred to as Neural Architecture Search (NAS), in designing neural network architectures has recently drawn considerable attention. In this work, we present an efficient NAS approach, named HM- NAS, that generalizes existing weight sharing based NAS approaches. Existing weight sharing based NAS approaches still adopt hand-designed heuristics to generate architecture candidates. As a consequence, the space of architecture candidates is constrained in a subset of all possible architectures, making the architecture search results sub-optimal. HM-NAS addresses this limitation via two innovations. First, HM-NAS incorporates a multi-level architecture encoding scheme to enable searching for more flexible network architectures. Second, it discards the hand-designed heuristics and incorporates a hierarchical masking scheme that automatically learns and determines the optimal architecture. Compared to state-of-the-art weight sharing based approaches, HM-NAS is able to achieve better architecture search performance and competitive model evaluation accuracy. Without the constraint imposed by the hand-designed heuristics, our searched networks contain more flexible and meaningful architectures that existing weight sharing based NAS approaches are not able to discover.

cs.LG↗
arXiv2019

Word-based Domain Adaptation for Neural Machine Translation

In this paper, we empirically investigate applying word-level weights to adapt neural machine translation to e-commerce domains, where small e-commerce datasets and large out-of-domain datasets are available. In order to mine in-domain like words in the out-of-domain datasets, we compute word weights by using a domain-specific and a non-domain-specific language model followed by smoothing and binary quantization. The baseline model is trained on mixed in-domain and out-of-domain datasets. Experimental results on English to Chinese e-commerce domain translation show that compared to continuing training without word weights, it improves MT quality by up to 2.11% BLEU absolute and 1.59% TER. We have also trained models using fine-tuning on the in-domain data. Pre-training a model with word weights improves fine-tuning up to 1.24% BLEU absolute and 1.64% TER, respectively.

cs.CL↗
arXiv2018

Social Bots for Online Public Health Interventions

According to the Center for Disease Control and Prevention, in the United States hundreds of thousands initiate smoking each year, and millions live with smoking-related dis- eases. Many tobacco users discuss their habits and preferences on social media. This work conceptualizes a framework for targeted health interventions to inform tobacco users about the consequences of tobacco use. We designed a Twitter bot named Notobot (short for No-Tobacco Bot) that leverages machine learning to identify users posting pro-tobacco tweets and select individualized interventions to address their interest in tobacco use. We searched the Twitter feed for tobacco-related keywords and phrases, and trained a convolutional neural network using over 4,000 tweets dichotomously manually labeled as either pro- tobacco or not pro-tobacco. This model achieves a 90% recall rate on the training set and 74% on test data. Users posting pro- tobacco tweets are matched with former smokers with similar interests who posted anti-tobacco tweets. Algorithmic matching, based on the power of peer influence, allows for the systematic delivery of personalized interventions based on real anti-tobacco tweets from former smokers. Experimental evaluation suggests that our system would perform well if deployed. This research offers opportunities for public health researchers to increase health awareness at scale. Future work entails deploying the fully operational Notobot system in a controlled experiment within a public health campaign.

cs.SI↗