scholarly journals Intrinsic RGB and multispectral images recovery by independent quadratic programming

2020 ◽  
Vol 6 ◽  
pp. e256
Author(s):  
Alexandre Krebs ◽  
Yannick Benezeth ◽  
Franck Marzani
Keyword(s):  
Quadratic Programming ◽  
Shape Recovery ◽  
State Of The Art ◽  
Model Parameters ◽  
Model Estimation ◽  
Multispectral Images ◽  
Single Image ◽  
Intrinsic Images ◽  
Previous Algorithm ◽  
Dichromatic Model

This work introduces a method to estimate reflectance, shading, and specularity from a single image. Reflectance, shading, and specularity are intrinsic images derived from the dichromatic model. Estimation of these intrinsic images has many applications in computer vision such as shape recovery, specularity removal, segmentation, or classification. The proposed method allows for recovering the dichromatic model parameters thanks to two independent quadratic programming steps. Compared to the state of the art in this domain, our approach has the advantage to address a complex inverse problem into two parallelizable optimization steps that are easy to solve and do not require learning. The proposed method is an extension of a previous algorithm that is rewritten to be numerically more stable, has better quantitative and qualitative results, and applies to multispectral images. The proposed method is assessed qualitatively and quantitatively on standard RGB and multispectral datasets.

EDPANs: Enhanced Dual Path Attention Networks for Single Image Super-Resolution

2021 ◽  
pp. 2150300
Author(s):  
Guoan Cheng ◽  
Ai Matsune ◽  
Huaijuan Zang ◽  
Toru Kurihara ◽  
Shu Zhan
Keyword(s):  
State Of The Art ◽  
Super Resolution ◽  
Single Image ◽  
Residual Network ◽  
Dense Structure ◽  
Attention Networks ◽  
Residual Structure ◽  
Quantitative Results ◽  
Image Super Resolution ◽  
Single Image Super Resolution

In this paper, we propose an enhanced dual path attention network (EDPAN) for image super-resolution. ResNet is good at implicitly reusing extracted features, DenseNet is good at exploring new features. Dual Path Network (DPN) combines ResNets and DenseNet to create a more accurate architecture than the straightforward one. We experimentally show that the residual network performs best when each block consists of two convolutions, and the dense network performs best when each micro-block consists of one convolution. Following these ideas, our EDPAN exploits the advantages of the residual structure and the dense structure. Besides, to deploy the computations for features more effectively, we introduce the attention mechanism into our EDPAN. Moreover, to relieve the parameters burden, we also utilize recursive learning to propose a lightweight model. In the experiments, we demonstrate the effectiveness and robustness of our proposed EDPAN on different degradation situations. The quantitative results and visualization comparison can sufficiently indicate that our EDPAN achieves favorable performance over the state-of-the-art frameworks.

scholarly journals Analysis of State-of-the-Art Spin-Transfer-Torque Nonvolatile Flip-Flops Considering Restore Yield in the Near/Sub-Threshold Voltage Region

Electronics ◽  
2020 ◽  
Vol 9 (12) ◽  
pp. 2118
Author(s):  
Gwang Hui Choi ◽  
Taehui Na
Keyword(s):  
Threshold Voltage ◽  
State Of The Art ◽  
Spin Transfer Torque ◽  
Spin Transfer ◽  
Model Parameters ◽  
Offset Cancellation ◽  
Battery Lifetime ◽  
Battery Capacity ◽  
Iot Devices ◽  
Voltage Region

Recently, the leakage power consumption of Internet of Things (IoT) devices has become a main issue to be tackled, due to the fact that the scaling of process technology increases the leakage current in the IoT devices having limited battery capacity, resulting in the reduction of battery lifetime. The most effective method to extend the battery lifetime is to shut-off the device during standby mode. For this reason, spin-transfer-torque magnetic-tunnel-junction (STT-MTJ) based nonvolatile flip-flop (NVFF) is being considered as a strong candidate to store the computing data. Since there is a risk that the MTJ resistance may change during the read operation (i.e., the read disturbance problem), NVFF should consider the read disturbance problem to satisfy reliable data restoration. To date, several NVFFs have been proposed. Even though they satisfy the target restore yield of 4σ, most of them do not take the read disturbance into account. Furthermore, several recently proposed NVFFs which focus on the offset-cancellation technique to improve the restore yield have obvious limitation with decreasing the supply voltage (VDD), because the offset-cancellation technique uses switch operation in the critical path that can exacerbate the restore yield in the near/sub-threshold region. In this regard, this paper analyzes state-of-the-art STT-MTJ based NVFFs with respect to the voltage region and provides insight that a simple circuit having no offset-cancellation technique could achieve a better restore yield in the near/sub-threshold voltage region. Monte–Carlo HSPICE simulation results, using industry-compatible 28 nm model parameters, show that in case of VDD of 0.6 V, complex NVFF circuits having offset tolerance characteristic have a better restore yield, whereas in case of VDD of 0.4 V with sizing up strategy, a simple NVFF circuit having no offset tolerance characteristic has a better restore yield.

scholarly journals Data-Efficient Sensor Upgrade Path Using Knowledge Distillation

Sensors ◽  
2021 ◽  
Vol 21 (19) ◽  
pp. 6523
Author(s):  
Pieter Van Van Molle ◽  
Cedric De De Boom ◽  
Tim Verbelen ◽  
Bert Vankeirsbilck ◽  
Jonas De De Vylder ◽  
...  
Keyword(s):  
Deep Neural Networks ◽  
State Of The Art ◽  
Original Data ◽  
Radar Data ◽  
Teacher Supervision ◽  
Multispectral Images ◽  
Test Set ◽  
Time To Market ◽  
Speed Up ◽  
Knowledge Distillation

Deep neural networks have achieved state-of-the-art performance in image classification. Due to this success, deep learning is now also being applied to other data modalities such as multispectral images, lidar and radar data. However, successfully training a deep neural network requires a large reddataset. Therefore, transitioning to a new sensor modality (e.g., from regular camera images to multispectral camera images) might result in a drop in performance, due to the limited availability of data in the new modality. This might hinder the adoption rate and time to market for new sensor technologies. In this paper, we present an approach to leverage the knowledge of a teacher network, that was trained using the original data modality, to improve the performance of a student network on a new data modality: a technique known in literature as knowledge distillation. By applying knowledge distillation to the problem of sensor transition, we can greatly speed up this process. We validate this approach using a multimodal version of the MNIST dataset. Especially when little data is available in the new modality (i.e., 10 images), training with additional teacher supervision results in increased performance, with the student network scoring a test set accuracy of 0.77, compared to an accuracy of 0.37 for the baseline. We also explore two extensions to the default method of knowledge distillation, which we evaluate on a multimodal version of the CIFAR-10 dataset: an annealing scheme for the hyperparameter α and selective knowledge distillation. Of these two, the first yields the best results. Choosing the optimal annealing scheme results in an increase in test set accuracy of 6%. Finally, we apply our method to the real-world use case of skin lesion classification.

scholarly journals Active Learning in the Geometric Block Model

2020 ◽  
Vol 34 (04) ◽  
pp. 3641-3648 ◽  
Author(s):  
Eli Chien ◽  
Antonia Tulino ◽  
Jaime Llorca
Keyword(s):  
Active Learning ◽  
Community Detection ◽  
Random Graphs ◽  
State Of The Art ◽  
Superior Performance ◽  
Model Parameters ◽  
Block Model ◽  
Stochastic Block Model ◽  
Synthetic Datasets ◽  
Community Detection Problems

The geometric block model is a recently proposed generative model for random graphs that is able to capture the inherent geometric properties of many community detection problems, providing more accurate characterizations of practical community structures compared with the popular stochastic block model. Galhotra et al. recently proposed a motif-counting algorithm for unsupervised community detection in the geometric block model that is proved to be near-optimal. They also characterized the regimes of the model parameters for which the proposed algorithm can achieve exact recovery. In this work, we initiate the study of active learning in the geometric block model. That is, we are interested in the problem of exactly recovering the community structure of random graphs following the geometric block model under arbitrary model parameters, by possibly querying the labels of a limited number of chosen nodes. We propose two active learning algorithms that combine the use of motif-counting with two different label query policies. Our main contribution is to show that sampling the labels of a vanishingly small fraction of nodes (sub-linear in the total number of nodes) is sufficient to achieve exact recovery in the regimes under which the state-of-the-art unsupervised method fails. We validate the superior performance of our algorithms via numerical simulations on both real and synthetic datasets.

scholarly journals AtLoc: Attention Guided Camera Localization

2020 ◽  
Vol 34 (06) ◽  
pp. 10393-10401
Author(s):  
Bing Wang ◽  
Changhao Chen ◽  
Chris Xiaoxuan Lu ◽  
Peijun Zhao ◽  
Niki Trigoni ◽  
...  
Keyword(s):  
Deep Learning ◽  
Experimental Evidence ◽  
State Of The Art ◽  
Source Code ◽  
Single Image ◽  
Saliency Maps ◽  
Camera Localization ◽  
Camera Pose ◽  
Dynamic Objects ◽  
Indoor And Outdoor

Deep learning has achieved impressive results in camera localization, but current single-image techniques typically suffer from a lack of robustness, leading to large outliers. To some extent, this has been tackled by sequential (multi-images) or geometry constraint approaches, which can learn to reject dynamic objects and illumination conditions to achieve better performance. In this work, we show that attention can be used to force the network to focus on more geometrically robust objects and features, achieving state-of-the-art performance in common benchmark, even if using only a single image as input. Extensive experimental evidence is provided through public indoor and outdoor datasets. Through visualization of the saliency maps, we demonstrate how the network learns to reject dynamic objects, yielding superior global camera pose regression performance. The source code is avaliable at https://github.com/BingCS/AtLoc.

scholarly journals Single Image Dehazing Using Sparse Contextual Representation

Atmosphere ◽  
2021 ◽  
Vol 12 (10) ◽  
pp. 1266
Author(s):  
Jing Qin ◽  
Liang Chen ◽  
Jian Xu ◽  
Wenqi Ren
Keyword(s):  
Sparse Representation ◽  
State Of The Art ◽  
Input Image ◽  
Single Image ◽  
Image Dehazing ◽  
Dark Channel ◽  
Local Patch ◽  
Novel Method ◽  
Contextual Representation ◽  
Single Image Dehazing

In this paper, we propose a novel method to remove haze from a single hazy input image based on the sparse representation. In our method, the sparse representation is proposed to be used as a contextual regularization tool, which can reduce the block artifacts and halos produced by only using dark channel prior without soft matting as the transmission is not always constant in a local patch. A novel way to use dictionary is proposed to smooth an image and generate the sharp dehazed result. Experimental results demonstrate that our proposed method performs favorably against the state-of-the-art dehazing methods and produces high-quality dehazed and vivid color results.

Shape from Profiles

2005 ◽  
Author(s):  
Roberto Cipolla ◽  
Kwan-Yee K. Wong
Keyword(s):  
Motion Estimation ◽  
State Of The Art ◽  
3D Models ◽  
Model Estimation ◽  
Model Reconstruction ◽  
The Creation ◽  
The Way

This chapter discusses profiles or outlines which are dominant features of images. Profiles can be extracted easily and reliably from the images and can provide information on the shape and motion of an object. Classical techniques for motion estimation and model reconstruction are highly dependent on point and line correspondences, hence they cannot be applied directly to profiles which are viewpoint dependent. The limitations of classical techniques paved the way for the creation of different sets of algorithms specific to profiles. In this chapter, the focus is on state-of-the-art algorithms for model reconstruction and model estimation from profiles. These new sets of algorithms are capable of reconstructing any kind of objects including smooth and textureless surfaces. They also render convincing 3D models, reinforcing the practicality of the algorithm.

scholarly journals Bayesian Planet Searches for the 10 cm/s Radial Velocity Era

2015 ◽  
Vol 11 (A29A) ◽  
pp. 205-207
Author(s):  
Philip C. Gregory
Keyword(s):  
Radial Velocity ◽  
State Of The Art ◽  
Simulated Data ◽  
Model Parameters ◽  
Data Sets ◽  
Stellar Activity ◽  
Bayesian Fusion ◽  
Multiple State ◽  
Simulated Data Sets ◽  
Apodization Function

AbstractA new apodized Keplerian model is proposed for the analysis of precision radial velocity (RV) data to model both planetary and stellar activity (SA) induced RV signals. A symmetrical Gaussian apodization function with unknown width and center can distinguish planetary signals from SA signals on the basis of the width of the apodization function. The general model for m apodized Keplerian signals also includes a linear regression term between RV and the stellar activity diagnostic In (R'hk), as well as an extra Gaussian noise term with unknown standard deviation. The model parameters are explored using a Bayesian fusion MCMC code. A differential version of the Generalized Lomb-Scargle periodogram provides an additional way of distinguishing SA signals and helps guide the choice of new periods. Sample results are reported for a recent international RV blind challenge which included multiple state of the art simulated data sets supported by a variety of stellar activity diagnostics.

scholarly journals Efficient Heterogeneous Collaborative Filtering without Negative Sampling for Recommendation

2020 ◽  
Vol 34 (01) ◽  
pp. 19-26 ◽  
Author(s):  
Chong Chen ◽  
Min Zhang ◽  
Yongfeng Zhang ◽  
Weizhi Ma ◽  
Yiqun Liu ◽  
...  
Keyword(s):  
Collaborative Filtering ◽  
Real World ◽  
Large Scale ◽  
State Of The Art ◽  
Heterogeneous Data ◽  
Model Parameters ◽  
Online Systems ◽  
Practical Applications ◽  
Real World Datasets ◽  
Primary Type

Recent studies on recommendation have largely focused on exploring state-of-the-art neural networks to improve the expressiveness of models, while typically apply the Negative Sampling (NS) strategy for efficient learning. Despite effectiveness, two important issues have not been well-considered in existing methods: 1) NS suffers from dramatic fluctuation, making sampling-based methods difficult to achieve the optimal ranking performance in practical applications; 2) although heterogeneous feedback (e.g., view, click, and purchase) is widespread in many online systems, most existing methods leverage only one primary type of user feedback such as purchase. In this work, we propose a novel non-sampling transfer learning solution, named Efficient Heterogeneous Collaborative Filtering (EHCF) for Top-N recommendation. It can not only model fine-grained user-item relations, but also efficiently learn model parameters from the whole heterogeneous data (including all unlabeled data) with a rather low time complexity. Extensive experiments on three real-world datasets show that EHCF significantly outperforms state-of-the-art recommendation methods in both traditional (single-behavior) and heterogeneous scenarios. Moreover, EHCF shows significant improvements in training efficiency, making it more applicable to real-world large-scale systems. Our implementation has been released 1 to facilitate further developments on efficient whole-data based neural methods.

scholarly journals Multiscale Single Image Dehazing Based on Adaptive Wavelet Fusion

2015 ◽  
Vol 2015 ◽  
pp. 1-14 ◽  
Author(s):  
Wei Wang ◽  
Wenhui Li ◽  
Qingji Guan ◽  
Miao Qi
Keyword(s):  
State Of The Art ◽  
Local Similarity ◽  
Fusion Method ◽  
Single Image ◽  
Image Dehazing ◽  
Adaptive Wavelet ◽  
Atmospheric Scattering ◽  
Dark Channel ◽  
Computer Vision Applications ◽  
Scene Depth

Removing the haze effects on images or videos is a challenging and meaningful task for image processing and computer vision applications. In this paper, we propose a multiscale fusion method to remove the haze from a single image. Based on the existing dark channel prior and optics theory, two atmospheric veils with different scales are first derived from the hazy image. Then, a novel and adaptive local similarity-based wavelet fusion method is proposed for preserving the significant scene depth property and avoiding blocky artifacts. Finally, the clear haze-free image is restored by solving the atmospheric scattering model. Experimental results demonstrate that the proposed method can yield comparative or even better results than several state-of-the-art methods by subjective and objective evaluations.

Sign in / Sign up

Export Citation Format

Share Document