arXiv subjects

Mei Chen

Publications and source records attributed to Mei Chen.

46 records · Page 3Linked to original sources

arXiv2021

Seismic Inverse Modeling Method based on Generative Adversarial Network

Seismic inverse modeling is a common method in reservoir prediction and it plays a vital role in the exploration and development of oil and gas. Conventional seismic inversion method is difficult to combine with complicated and abstract knowledge on geological mode and its uncertainty is difficult to be assessed. The paper proposes an inversion modeling method based on GAN consistent with geology, well logs, seismic data. GAN is a the most promising generation model algorithm that extracts spatial structure and abstract features of training images. The trained GAN can reproduce the models with specific mode. In our test, 1000 models were generated in 1 second. Based on the trained GAN after assessment, the optimal result of models can be calculated through Bayesian inversion frame. Results show that inversion models conform to observation data and have a low uncertainty under the premise of fast generation. This seismic inverse modeling method increases the efficiency and quality of inversion iteration. It is worthy of studying and applying in fusion of seismic data and geological knowledge.

stat.ML↗

arXiv2021

Revisiting Dynamic Convolution via Matrix Decomposition

Recent research in dynamic convolution shows substantial performance boost for efficient CNNs, due to the adaptive aggregation of K static convolution kernels. It has two limitations: (a) it increases the number of convolutional weights by K-times, and (b) the joint optimization of dynamic attention and static convolution kernels is challenging. In this paper, we revisit it from a new perspective of matrix decomposition and reveal the key issue is that dynamic convolution applies dynamic attention over channel groups after projecting into a higher dimensional latent space. To address this issue, we propose dynamic channel fusion to replace dynamic attention over channel groups. Dynamic channel fusion not only enables significant dimension reduction of the latent space, but also mitigates the joint optimization difficulty. As a result, our method is easier to train and requires significantly fewer parameters without sacrificing accuracy. Source code is at https://github.com/liyunsheng13/dcd.

cs.CV↗

arXiv2020

RasterNet: Modeling Free-Flow Speed using LiDAR and Overhead Imagery

Roadway free-flow speed captures the typical vehicle speed in low traffic conditions. Modeling free-flow speed is an important problem in transportation engineering with applications to a variety of design, operation, planning, and policy decisions of highway systems. Unfortunately, collecting large-scale historical traffic speed data is expensive and time consuming. Traditional approaches for estimating free-flow speed use geometric properties of the underlying road segment, such as grade, curvature, lane width, lateral clearance and access point density, but for many roads such features are unavailable. We propose a fully automated approach, RasterNet, for estimating free-flow speed without the need for explicit geometric features. RasterNet is a neural network that fuses large-scale overhead imagery and aerial LiDAR point clouds using a geospatially consistent raster structure. To support training and evaluation, we introduce a novel dataset combining free-flow speeds of road segments, overhead imagery, and LiDAR point clouds across the state of Kentucky. Our method achieves state-of-the-art results on a benchmark dataset.

cs.CV↗

arXiv2020

HyperSTAR: Task-Aware Hyperparameters for Deep Networks

While deep neural networks excel in solving visual recognition tasks, they require significant effort to find hyperparameters that make them work optimally. Hyperparameter Optimization (HPO) approaches have automated the process of finding good hyperparameters but they do not adapt to a given task (task-agnostic), making them computationally inefficient. To reduce HPO time, we present HyperSTAR (System for Task Aware Hyperparameter Recommendation), a task-aware method to warm-start HPO for deep neural networks. HyperSTAR ranks and recommends hyperparameters by predicting their performance conditioned on a joint dataset-hyperparameter space. It learns a dataset (task) representation along with the performance predictor directly from raw images in an end-to-end fashion. The recommendations, when integrated with an existing HPO method, make it task-aware and significantly reduce the time to achieve optimal performance. We conduct extensive experiments on 10 publicly available large-scale image classification datasets over two different network architectures, validating that HyperSTAR evaluates 50% less configurations to achieve the best performance compared to existing methods. We further demonstrate that HyperSTAR makes Hyperband (HB) task-aware, achieving the optimal accuracy in just 25% of the budget required by both vanilla HB and Bayesian Optimized HB~(BOHB).

cs.CV↗

arXiv2020

Unsupervised Instance Segmentation in Microscopy Images via Panoptic Domain Adaptation and Task Re-weighting

Unsupervised domain adaptation (UDA) for nuclei instance segmentation is important for digital pathology, as it alleviates the burden of labor-intensive annotation and domain shift across datasets. In this work, we propose a Cycle Consistency Panoptic Domain Adaptive Mask R-CNN (CyC-PDAM) architecture for unsupervised nuclei segmentation in histopathology images, by learning from fluorescence microscopy images. More specifically, we first propose a nuclei inpainting mechanism to remove the auxiliary generated objects in the synthesized images. Secondly, a semantic branch with a domain discriminator is designed to achieve panoptic-level domain adaptation. Thirdly, in order to avoid the influence of the source-biased features, we propose a task re-weighting mechanism to dynamically add trade-off weights for the task-specific loss functions. Experimental results on three datasets indicate that our proposed method outperforms state-of-the-art UDA methods significantly, and demonstrates a similar performance as fully supervised methods.

cs.CV↗

arXiv2019

Region and Object based Panoptic Image Synthesis through Conditional GANs

Image-to-image translation is significant to many computer vision and machine learning tasks such as image synthesis and video synthesis. It has primary applications in the graphics editing and animation industries. With the development of generative adversarial networks, a lot of attention has been drawn to image-to-image translation tasks. In this paper, we propose and investigate a novel task named as panoptic-level image-to-image translation and a naive baseline of solving this task. Panoptic-level image translation extends the current image translation task to two separate objectives of semantic style translation (adjust the style of objects to that of different domains) and instance transfiguration (swap between different types of objects). The proposed task generates an image from a complete and detailed panoptic perspective which can enrich the context of real-world vision synthesis. Our contribution consists of the proposal of a significant task worth investigating and a naive baseline of solving it. The proposed baseline consists of the multiple instances sequential translation and semantic-level translation with domain-invariant content code.

cs.CV↗

arXiv2019

FARSA: Fully Automated Roadway Safety Assessment

This paper addresses the task of road safety assessment. An emerging approach for conducting such assessments in the United States is through the US Road Assessment Program (usRAP), which rates roads from highest risk (1 star) to lowest (5 stars). Obtaining these ratings requires manual, fine-grained labeling of roadway features in street-level panoramas, a slow and costly process. We propose to automate this process using a deep convolutional neural network that directly estimates the star rating from a street-level panorama, requiring milliseconds per image at test time. Our network also estimates many other road-level attributes, including curvature, roadside hazards, and the type of median. To support this, we incorporate task-specific attention layers so the network can focus on the panorama regions that are most useful for a particular task. We evaluated our approach on a large dataset of real-world images from two US states. We found that incorporating additional tasks, and using a semi-supervised training approach, significantly reduced overfitting problems, allowed us to optimize more layers of the network, and resulted in higher accuracy.

cs.CV↗

arXiv2018

Enhancing Stratified Graph Sampling Algorithms based on Approximate Degree Distribution

Sampling technique has become one of the recent research focuses in the graph-related fields. Most of the existing graph sampling algorithms tend to sample the high degree or low degree nodes in the complex networks because of the characteristic of scale-free. Scale-free means that degrees of different nodes are subject to a power law distribution. So, there is a significant difference in the degrees between the overall sampling nodes. In this paper, we propose an idea of approximate degree distribution and devise a stratified strategy using it in the complex networks. We also develop two graph sampling algorithms combining the node selection method with the stratified strategy. The experimental results show that our sampling algorithms preserve several properties of different graphs and behave more accurately than other algorithms. Further, we prove the proposed algorithms are superior to the off-the-shelf algorithms in terms of the unbiasedness of the degrees and more efficient than state-of-the-art FFS and ES-i algorithms.

cs.SI↗

arXiv2005

A Capacity Achieving and Low Complexity Multilevel Coding Scheme for ISI Channels

We propose a computationally efficient multilevel coding scheme to achieve the capacity of an ISI channel using layers of binary inputs. The transmitter employs multilevel coding with linear mapping. The receiver uses multistage decoding where each stage performs a separate linear minimum mean square error (LMMSE) equalization and decoding. The optimality of the scheme is due to the fact that the LMMSE equalizer is information lossless in an ISI channel when signal to noise ratio is sufficiently low. The computational complexity is low and scales linearly with the length of the channel impulse response and the number of layers. The decoder at each layer sees an equivalent AWGN channel, which makes coding straightforward.

cs.IT↗

arXiv2005

Trellis Pruning for Peak-to-Average Power Ratio Reduction

This paper introduces a new trellis pruning method which uses nonlinear convolutional coding for peak-to-average power ratio (PAPR) reduction of filtered QPSK and 16-QAM modulations. The Nyquist filter is viewed as a convolutional encoder that controls the analog waveforms of the filter output directly. Pruning some edges of the encoder trellis can effectively reduce the PAPR. The only tradeoff is a slightly lower channel capacity and increased complexity. The paper presents simulation results of the pruning action and the resulting PAPR, and also discusses the decoding algorithm and the capacity of the filtered and pruned QPSK and 16-QAM modulations on the AWGN channel. Simulation results show that the pruning method reduces the PAPR significantly without much damage to capacity.

cs.IT↗