April 2, 2020

2999 words 15 mins read

Paper Group ANR 128

Paper Group ANR 128

Time-aware Large Kernel Convolutions. Symmetry Detection of Occluded Point Cloud Using Deep Learning. Deep Blind Video Super-resolution. ROSE: Real One-Stage Effort to Detect the Fingerprint Singular Point Based on Multi-scale Spatial Attention. EgoMap: Projective mapping and structured egocentric memory for Deep RL. Relevant-features based Auxilia …

Time-aware Large Kernel Convolutions

Title Time-aware Large Kernel Convolutions
Authors Vasileios Lioutas, Yuhong Guo
Abstract To date, most state-of-the-art sequence modelling architectures use attention to build generative models for language based tasks. Some of these models use all the available sequence tokens to generate an attention distribution which results in time complexity of $O(n^2)$. Alternatively, they utilize depthwise convolutions with softmax normalized kernels of size $k$ acting as a limited-window self-attention, resulting in time complexity of $O(k{\cdot}n)$. In this paper, we introduce Time-aware Large Kernel (TaLK) Convolutions, a novel adaptive convolution operation that learns to predict the size of a summation kernel instead of using the fixed-sized kernel matrix. This method yields a time complexity of $O(n)$, effectively making the sequence encoding process linear to the number of tokens. We evaluate the proposed method on large-scale standard machine translation and language modelling datasets and show that TaLK Convolutions constitute an efficient improvement over other attention/convolution based approaches.
Tasks Language Modelling, Machine Translation
Published 2020-02-08
URL https://arxiv.org/abs/2002.03184v1
PDF https://arxiv.org/pdf/2002.03184v1.pdf
PWC https://paperswithcode.com/paper/time-aware-large-kernel-convolutions

Symmetry Detection of Occluded Point Cloud Using Deep Learning

Title Symmetry Detection of Occluded Point Cloud Using Deep Learning
Authors Zhelun Wu, Hongyan Jiang, Siyun He
Abstract Symmetry detection has been a classical problem in computer graphics, many of which using traditional geometric methods. In recent years, however, we have witnessed the arising deep learning changed the landscape of computer graphics. In this paper, we aim to solve the symmetry detection of the occluded point cloud in a deep-learning fashion. To the best of our knowledge, we are the first to utilize deep learning to tackle such a problem. In such a deep learning framework, double supervisions: points on the symmetry plane and normal vectors are employed to help us pinpoint the symmetry plane. We conducted experiments on the YCB- video dataset and demonstrate the efficacy of our method.
Published 2020-03-14
URL https://arxiv.org/abs/2003.06520v1
PDF https://arxiv.org/pdf/2003.06520v1.pdf
PWC https://paperswithcode.com/paper/symmetry-detection-of-occluded-point-cloud

Deep Blind Video Super-resolution

Title Deep Blind Video Super-resolution
Authors Jinshan Pan, Songsheng Cheng, Jiawei Zhang, Jinhui Tang
Abstract Existing video super-resolution (SR) algorithms usually assume that the blur kernels in the degradation process are known and do not model the blur kernels in the restoration. However, this assumption does not hold for video SR and usually leads to over-smoothed super-resolved images. In this paper, we propose a deep convolutional neural network (CNN) model to solve video SR by a blur kernel modeling approach. The proposed deep CNN model consists of motion blur estimation, motion estimation, and latent image restoration modules. The motion blur estimation module is used to provide reliable blur kernels. With the estimated blur kernel, we develop an image deconvolution method based on the image formation model of video SR to generate intermediate latent images so that some sharp image contents can be restored well. However, the generated intermediate latent images may contain artifacts. To generate high-quality images, we use the motion estimation module to explore the information from adjacent frames, where the motion estimation can constrain the deep CNN model for better image restoration. We show that the proposed algorithm is able to generate clearer images with finer structural details. Extensive experimental results show that the proposed algorithm performs favorably against state-of-the-art methods.
Tasks Image Deconvolution, Image Restoration, Motion Estimation, Super-Resolution, Video Super-Resolution
Published 2020-03-10
URL https://arxiv.org/abs/2003.04716v1
PDF https://arxiv.org/pdf/2003.04716v1.pdf
PWC https://paperswithcode.com/paper/deep-blind-video-super-resolution

ROSE: Real One-Stage Effort to Detect the Fingerprint Singular Point Based on Multi-scale Spatial Attention

Title ROSE: Real One-Stage Effort to Detect the Fingerprint Singular Point Based on Multi-scale Spatial Attention
Authors Liaojun Pang, Jiong Chen, Fei Guo, Zhicheng Cao, Heng Zhao
Abstract Detecting the singular point accurately and efficiently is one of the most important tasks for fingerprint recognition. In recent years, deep learning has been gradually used in the fingerprint singular point detection. However, current deep learning-based singular point detection methods are either two-stage or multi-stage, which makes them time-consuming. More importantly, their detection accuracy is yet unsatisfactory, especially in the case of the low-quality fingerprint. In this paper, we make a Real One-Stage Effort to detect fingerprint singular points more accurately and efficiently, and therefore we name the proposed algorithm ROSE for short, in which the multi-scale spatial attention, the Gaussian heatmap and the variant of focal loss are applied together to achieve a higher detection rate. Experimental results on the datasets FVC2002 DB1 and NIST SD4 show that our ROSE outperforms the state-of-art algorithms in terms of detection rate, false alarm rate and detection speed.
Published 2020-03-09
URL https://arxiv.org/abs/2003.03918v1
PDF https://arxiv.org/pdf/2003.03918v1.pdf
PWC https://paperswithcode.com/paper/rose-real-one-stage-effort-to-detect-the

EgoMap: Projective mapping and structured egocentric memory for Deep RL

Title EgoMap: Projective mapping and structured egocentric memory for Deep RL
Authors Edward Beeching, Christian Wolf, Jilles Dibangoye, Olivier Simonin
Abstract Tasks involving localization, memorization and planning in partially observable 3D environments are an ongoing challenge in Deep Reinforcement Learning. We present EgoMap, a spatially structured neural memory architecture. EgoMap augments a deep reinforcement learning agent’s performance in 3D environments on challenging tasks with multi-step objectives. The EgoMap architecture incorporates several inductive biases including a differentiable inverse projection of CNN feature vectors onto a top-down spatially structured map. The map is updated with ego-motion measurements through a differentiable affine transform. We show this architecture outperforms both standard recurrent agents and state of the art agents with structured memory. We demonstrate that incorporating these inductive biases into an agent’s architecture allows for stable training with reward alone, circumventing the expense of acquiring and labelling expert trajectories. A detailed ablation study demonstrates the impact of key aspects of the architecture and through extensive qualitative analysis, we show how the agent exploits its structured internal memory to achieve higher performance.
Published 2020-01-24
URL https://arxiv.org/abs/2002.02286v2
PDF https://arxiv.org/pdf/2002.02286v2.pdf
PWC https://paperswithcode.com/paper/egomap-projective-mapping-and-structured-1

Relevant-features based Auxiliary Cells for Energy Efficient Detection of Natural Errors

Title Relevant-features based Auxiliary Cells for Energy Efficient Detection of Natural Errors
Authors Sai Aparna Aketi, Priyadarshini Panda, Kaushik Roy
Abstract Deep neural networks have demonstrated state-of-the-art performance on many classification tasks. However, they have no inherent capability to recognize when their predictions are wrong. There have been several efforts in the recent past to detect natural errors but the suggested mechanisms pose additional energy requirements. To address this issue, we propose an ensemble of classifiers at hidden layers to enable energy efficient detection of natural errors. In particular, we append Relevant-features based Auxiliary Cells (RACs) which are class specific binary linear classifiers trained on relevant features. The consensus of RACs is used to detect natural errors. Based on combined confidence of RACs, classification can be terminated early, thereby resulting in energy efficient detection. We demonstrate the effectiveness of our technique on various image classification datasets such as CIFAR-10, CIFAR-100 and Tiny-ImageNet.
Tasks Image Classification
Published 2020-02-25
URL https://arxiv.org/abs/2002.11052v2
PDF https://arxiv.org/pdf/2002.11052v2.pdf
PWC https://paperswithcode.com/paper/relevant-features-based-auxiliary-cells-for-1

Differential Network Analysis: A Statistical Perspective

Title Differential Network Analysis: A Statistical Perspective
Authors Ali Shojaie
Abstract Networks effectively capture interactions among components of complex systems, and have thus become a mainstay in many scientific disciplines. Growing evidence, especially from biology, suggest that networks undergo changes over time, and in response to external stimuli. In biology and medicine, these changes have been found to be predictive of complex diseases. They have also been used to gain insight into mechanisms of disease initiation and progression. Primarily motivated by biological applications, this article provides a review of recent statistical machine learning methods for inferring networks and identifying changes in their structures.
Published 2020-03-09
URL https://arxiv.org/abs/2003.04235v1
PDF https://arxiv.org/pdf/2003.04235v1.pdf
PWC https://paperswithcode.com/paper/differential-network-analysis-a-statistical

Engaging Users through Social Media in Public Libraries

Title Engaging Users through Social Media in Public Libraries
Authors Hongbo Zou, Hsuanwei Michelle Chen, Sharmistha Dey
Abstract The participatory library is an emerging concept which refers to the idea that an integrated library system must allow users to take part in core functions of the library rather than engaging on the periphery. To embrace the participatory idea, libraries have employed many technologies, such as social media to help them build participatory services and engage users. To help librarians understand the impact of emerging technologies on a participatory service building, this paper takes social media as an example to explore how to use different engagement strategies that social media provides to engage more users. This paper provides three major contributions to the library system. The libraries can use the resultant engagement strategies to engage its users. Additionally, the best-fit strategy can be inferred and designed based on the preferences of users. Lastly, the preferences of users can be understood based on data analysis of social media. Three such contributions put together to fully address the proposed research question of how to use different engagement strategies on social media to build participatory library services and better engage more users visiting the library?
Published 2020-02-25
URL https://arxiv.org/abs/2003.04204v1
PDF https://arxiv.org/pdf/2003.04204v1.pdf
PWC https://paperswithcode.com/paper/engaging-users-through-social-media-in-public

CURE Dataset: Ladder Networks for Audio Event Classification

Title CURE Dataset: Ladder Networks for Audio Event Classification
Authors Harishchandra Dubey, Dimitra Emmanouilidou, Ivan J. Tashev
Abstract Audio event classification is an important task for several applications such as surveillance, audio, video and multimedia retrieval etc. There are approximately 3M people with hearing loss who can’t perceive events happening around them. This paper establishes the CURE dataset which contains curated set of specific audio events most relevant for people with hearing loss. We propose a ladder network based audio event classifier that utilizes 5s sound recordings derived from the Freesound project. We adopted the state-of-the-art convolutional neural network (CNN) embeddings as audio features for this task. We also investigate extreme learning machine (ELM) for event classification. In this study, proposed classifiers are compared with support vector machine (SVM) baseline. We propose signal and feature normalization that aims to reduce the mismatch between different recordings scenarios. Firstly, CNN is trained on weakly labeled Audioset data. Next, the pre-trained model is adopted as feature extractor for proposed CURE corpus. We incorporate ESC-50 dataset as second evaluation set. Results and discussions validate the superiority of Ladder network over ELM and SVM classifier in terms of robustness and increased classification accuracy. While Ladder network is robust to data mismatches, simpler SVM and ELM classifiers are sensitive to such mismatches, where the proposed normalization techniques can play an important role. Experimental studies with ESC-50 and CURE corpora elucidate the differences in dataset complexity and robustness offered by proposed approaches.
Published 2020-01-12
URL https://arxiv.org/abs/2001.03896v1
PDF https://arxiv.org/pdf/2001.03896v1.pdf
PWC https://paperswithcode.com/paper/cure-dataset-ladder-networks-for-audio-event

Utilizing Differential Evolution into optimizing targeted cancer treatments

Title Utilizing Differential Evolution into optimizing targeted cancer treatments
Authors Michail-Antisthenis Tsompanas, Larry Bull, Andrew Adamatzky, Igor Balaz
Abstract Working towards the development of an evolvable cancer treatment simulator, the investigation of Differential Evolution was considered, motivated by the high efficiency of variations of this technique in real-valued problems. A basic DE algorithm, namely “DE/rand/1” was used to optimize the simulated design of a targeted drug delivery system for tumor treatment on PhysiCell simulator. The suggested approach proved to be more efficient than a standard genetic algorithm, which was not able to escape local minima after a predefined number of generations. The key attribute of DE that enables it to outperform standard EAs, is the fact that it keeps the diversity of the population high, throughout all the generations. This work will be incorporated with ongoing research in a more wide applicability platform that will design, develop and evaluate targeted drug delivery systems aiming cancer tumours.
Published 2020-03-21
URL https://arxiv.org/abs/2003.11623v1
PDF https://arxiv.org/pdf/2003.11623v1.pdf
PWC https://paperswithcode.com/paper/utilizing-differential-evolution-into
Title PHS: A Toolbox for Parallel Hyperparameter Search
Authors Peter Michael Habelitz, Janis Keuper
Abstract We introduce an open source python framework named PHS - Parallel Hyperparameter Search to enable hyperparameter optimization on numerous compute instances of any arbitrary python function. This is achieved with minimal modifications inside the target function. Possible applications appear in expensive to evaluate numerical computations which strongly depend on hyperparameters such as machine learning. Bayesian optimization is chosen as a sample efficient method to propose the next query set of parameters.
Tasks Hyperparameter Optimization
Published 2020-02-26
URL https://arxiv.org/abs/2002.11429v2
PDF https://arxiv.org/pdf/2002.11429v2.pdf
PWC https://paperswithcode.com/paper/phs-a-toolbox-for-parellel-hyperparameter

PairNets: Novel Fast Shallow Artificial Neural Networks on Partitioned Subspaces

Title PairNets: Novel Fast Shallow Artificial Neural Networks on Partitioned Subspaces
Authors Luna M. Zhang
Abstract Traditionally, an artificial neural network (ANN) is trained slowly by a gradient descent algorithm such as the backpropagation algorithm since a large number of hyperparameters of the ANN need to be fine-tuned with many training epochs. To highly speed up training, we created a novel shallow 4-layer ANN called “Pairwise Neural Network” (“PairNet”) with high-speed hyperparameter optimization. In addition, a value of each input is partitioned into multiple intervals, and then an n-dimensional space is partitioned into M n-dimensional subspaces. M local PairNets are built in M partitioned local n-dimensional subspaces. A local PairNet is trained very quickly with only one epoch since its hyperparameters are directly optimized one-time via simply solving a system of linear equations by using the multivariate least squares fitting method. Simulation results for three regression problems indicated that the PairNet achieved much higher speeds and lower average testing mean squared errors (MSEs) for the three cases, and lower average training MSEs for two cases than the traditional ANNs. A significant future work is to develop better and faster optimization algorithms based on intelligent methods and parallel computing methods to optimize both partitioned subspaces and hyperparameters to build the fast and effective PairNets for applications in big data mining and real-time machine learning.
Tasks Hyperparameter Optimization
Published 2020-01-24
URL https://arxiv.org/abs/2001.08886v1
PDF https://arxiv.org/pdf/2001.08886v1.pdf
PWC https://paperswithcode.com/paper/pairnets-novel-fast-shallow-artificial-neural

DiffNet++: A Neural Influence and Interest Diffusion Network for Social Recommendation

Title DiffNet++: A Neural Influence and Interest Diffusion Network for Social Recommendation
Authors Le Wu, Junwei Li, Peijie Sun, Richang Hong, Yong Ge, Meng Wang
Abstract Social recommendation has emerged to leverage social connections among users for predicting users’ unknown preferences, which could alleviate the data sparsity issue in collaborative filtering based recommendation. Early approaches relied on utilizing each user’s first-order social neighbors’ interests for better user modeling and failed to model the social influence diffusion process from the global social network structure. Recently, we propose a preliminary work of a neural influence diffusion network (i.e., DiffNet) for social recommendation (Diffnet), which models the recursive social diffusion process to capture the higher-order relationships for each user. However, we argue that, as users play a central role in both user-user social network and user-item interest network, only modeling the influence diffusion process in the social network would neglect the users’ latent collaborative interests in the user-item interest network. In this paper, we propose DiffNet++, an improved algorithm of DiffNet that models the neural influence diffusion and interest diffusion in a unified framework. By reformulating the social recommendation as a heterogeneous graph with social network and interest network as input, DiffNet++ advances DiffNet by injecting these two network information for user embedding learning at the same time. This is achieved by iteratively aggregating each user’s embedding from three aspects: the user’s previous embedding, the influence aggregation of social neighbors from the social network, and the interest aggregation of item neighbors from the user-item interest network. Furthermore, we design a multi-level attention network that learns how to attentively aggregate user embeddings from these three aspects. Finally, extensive experimental results on two real-world datasets clearly show the effectiveness of our proposed model.
Published 2020-01-15
URL https://arxiv.org/abs/2002.00844v3
PDF https://arxiv.org/pdf/2002.00844v3.pdf
PWC https://paperswithcode.com/paper/diffnet-a-neural-influence-and-interest

A Novel AI-enabled Framework to Diagnose Coronavirus COVID 19 using Smartphone Embedded Sensors: Design Study

Title A Novel AI-enabled Framework to Diagnose Coronavirus COVID 19 using Smartphone Embedded Sensors: Design Study
Authors Halgurd S. Maghdid, Kayhan Zrar Ghafoor, Ali Safaa Sadiq, Kevin Curran, Khaled Rabie
Abstract Coronaviruses are a famous family of viruses that causes illness in human or animals. The new type of corona virus COVID-19 disease was firstly discovered in Wuhan-China. However, recently, the virus has been widely spread in most of the world countries and is reported as a pandemic. Further, nowadays, all the world countries are striving to control the coronavirus disease COVID-19. There are many mechanisms to detect the coronavirus disease COVID-19 including clinical analysis of chest CT scan images and blood test results. The confirmed COVID-19 patient manifests as fever, tiredness, and dry cough. Particularly, several techniques can be used to detect the initial results of the virus such as medical detection Kits. However, such devices are incurring huge cost and it takes time to install them and use. Therefore, in this paper, a new framework is proposed to detect coronavirus disease COVID-19 using onboard smartphone sensors. The proposal provides a low-cost solution, since most of the radiologists have already held smartphones for different daily-purposes. People can use the framework on their smartphones for the virus detection purpose. Nowadays, smartphones are powerful with existing computation-rich processors, memory space, and large number of sensors including cameras, microphone, temperature sensor, inertial sensors, proximity, colour-sensor, humidity-sensor, and wireless chipsets/sensors. The designed Artificial Intelligence (AI) enabled framework reads the smartphone sensors signal measurements to predict the grade of severity of the pneumonia as well as predicting the result of the disease.
Published 2020-03-16
URL https://arxiv.org/abs/2003.07434v1
PDF https://arxiv.org/pdf/2003.07434v1.pdf
PWC https://paperswithcode.com/paper/a-novel-ai-enabled-framework-to-diagnose

Memorizing Gaussians with no over-parameterizaion via gradient decent on neural networks

Title Memorizing Gaussians with no over-parameterizaion via gradient decent on neural networks
Authors Amit Daniely
Abstract We prove that a single step of gradient decent over depth two network, with $q$ hidden neurons, starting from orthogonal initialization, can memorize $\Omega\left(\frac{dq}{\log^4(d)}\right)$ independent and randomly labeled Gaussians in $\mathbb{R}^d$. The result is valid for a large class of activation functions, which includes the absolute value.
Published 2020-03-28
URL https://arxiv.org/abs/2003.12895v1
PDF https://arxiv.org/pdf/2003.12895v1.pdf
PWC https://paperswithcode.com/paper/memorizing-gaussians-with-no-over
comments powered by Disqus