Technical Report for Valence Arousal Estimation on Affwild2 Dataset

In this work, we describe our method for tackling the valence-arousalestimation challenge from ABAW FG-2020 Competition . The competition organizers provide an in-the-wild Aff-Wild2 dataset for participants to analyze affectivebehavior in real-life settings . We use MIMAMO Net to achieve information about micro-motion and macro-motion for improving videoemotion recognition .…

Secure Virtual Mobile Small Cells A Stepping Stone Towards 6G

As 5th Generation research reaches the twilight, the research community must look beyond 5G and look towards the 2030 connectivity landscape, namely 6G . Virtual mobile small cells (MSC) are created on demand and can harness radio and networking capability locally reducing protocol signaling latency and overhead .…

Architecture of a Flexible and Cost Effective Remote Code Execution Engine

A cloud-based web service for remote code execution, that is easily extensible to support any number of programming languages and libraries . The service provides a fast, reproduciblesolution for small software experiments and is amenable to collaboration in a workplace (via sharable permalinks) The service is designed as a distributedsystem to reliably support a large number of users, and efficiently managecloud-hosting costs with predictive auto-scaling while minimizing SLAviolations .…

Multiphysics model reduction of thermomechanical vibration in a state space formulation

The aim of this work is to propose a new multiphysics mode synthesis (MMS) The present thermomechanical model is based on a state-space formulation, which consists of displacement,velocity, and temperature shift . The proposed MMS strategy can also be also applied to accelerate thecomputational speed by using independent parallel solvers .…

Area Rate Efficiency in Molecular Communications

We consider a multiuser diffusion-based molecular communication (MC) system . Multiple spatially distributed transmitter (TX)-receiver (RX) pairs establish point-to-point communication links employing the same type of signaling molecules . We propose a new performance metric, which we referto as area rate efficiency (ARE), that captures the tradeoff between the userdensity and IUI .…

A Review of Confidentiality Threats Against Embedded Neural Network Models

Machine Learning (ML) algorithms, especially Deep NeuralNetwork (DNN) models, become a widely accepted standard in many domains moreparticularly IoT-based systems . DNN models reach impressive performances in several sensitive fields such as medical diagnosis, smart transport or securitythreat detection, and represent a valuable piece of Intellectual Property .…

The distributed dual ascent algorithm is robust to asynchrony

The distributed dual ascent is an established algorithm to solve stronglyconvex multi-agent optimization problems with separable cost functions . In this paper, we study its asynchronouscounterpart . We assume that each agent only relies on the outdated information received from some neighbors .…

Dual Cross Central Difference Network for Face Anti Spoofing

Face anti-spoofing (FAS) plays a vital role in securing face recognitionsystems . Central difference convolution (CDC) has shown its excellent capacity for the FAS task via leveraging local gradientfeatures . However, aggregating central difference clues from allneighbors/directions simultaneously makes CDC redundant and sub-optimized .…

Neural Weighted A Learning Graph Costs and Heuristics with Differentiable Anytime A

We propose Neural Weighted A* a differentiable planner able to produce improved representations of planar maps asgraph costs and heuristics . Training occurs end-to-end on raw images and direct supervision on planning examples . We outperform similar architectures in planning accuracy and efficiency, and can trade offplanning accuracy for efficiency at run-time, using a single, real-valuedparameter .…

Switching 3 edge colorings of cubic graphs

The chromatic index of a cubic graph is either 3 or 4 . Edge-Kempe switching can be used to transform edge-colorings of cubic graphs . It is further connected to cycle switching of Steiner triple systems, for example, for which an improvement of the classification algorithm is presented .…

Continuous indetermination and average likelihood minimization

Authors transpose a discrete notion of indetermination coupling in thecase of continuous probabilities . They show that this coupling, expressed ondensities, cannot be captured by a specific copula which acts on cumulativedistribution functions without a high dependence on the margins .…

Semantic Modeling for Food Recommendation Explanations

FoodExplanation Ontology (FEO) provides a formalism for modeling explanationsto users for food-related recommendations . FEO models food recommendations,using concepts from the explanation domain to create responses to userquestions about food recommendations they receive from AI systems such as personalized knowledge base question answering systems .…

HASCO Towards Agile HArdware and Software CO design for Tensor Computation

Tensor computations overwhelm traditional general-purpose computing devices . They callfor a holistic solution composed of both hardware acceleration and softwaremapping . Hardware/software (HW/SW) co-design optimizes the hardware andsoftware in concert and produces high-quality solutions . Hasco achieves a 1.25X to 1.44Xlatency reduction through HW/SW co-Design compared with developing the hardwareand software separately .…

Deterministic Rounding of Dynamic Fractional Matchings

We present a framework for deterministically rounding a dynamic fractional matching algorithm . This is the first dynamic matching algorithm that works on general graphs by using an algorithm for low-arboricity graphs as ablack-box subroutine . Our rounding scheme works by maintaining a good {\em matching-sparsifier} with bounded arboricity, and then applying the algorithm of Peleg and Solomon[SODA’16] to maintain a near-optimal matching in this low arboric graph .…

Orienting Point Clouds with Dipole Propagation

Establishing a consistent normal orientation for point clouds is anotoriously difficult problem in geometry processing . The normal direction of a point is a function of the local surface neighborhood; yet, point clouds do not disclosethe full underlying surface structure .…

LAFFNet A Lightweight Adaptive Feature Fusion Network for Underwater Image Enhancement

There are many deep-learning-based methods with impressive performance for underwater image enhancement, but their memory and model parameter costs are hindrances in practical application . Toaddress this issue, we propose a lightweight adaptive feature fusion network(LAFFNet) The model is the encoder-decoder model with multiple adaptivefeature fusion (AAF) modules .…

Alternate Model Growth and Pruning for Efficient Training of Recommendation Systems

Model pruning is an effective technique to reduce computation overhead for deep neural networks . Modern recommendation systems are thirsty for model capacity due to the demand for handling big data . Pruning a recommendation model at scale results in a smaller model capacity and lower accuracy .…

Broadly Applicable Targeted Data Sample Omission Attacks

We introduce a novel clean-label targeted poisoning attack on learningmechanisms . We show that, with a low attack budget, our attack’s success rate is above 80%, and in some cases 100%, for white-box learning . We demonstrate the effectiveness of omission attacks against a large variety of learners including Deep learning, SVM and decisiontrees .…

Hadamard matrices in 0 1 presentation and an algorithm for generating them

Hadamard matrices are square $n\times n$ matrices whose entries are ones and minus ones . Matrices are orthogonal to each other with respect to the standard scalar product in $\Bbb R^n$. Each Hadamar matrix can be transformed to a matrix whose entries were zeros and ones .…

Playing Stochastically in Weighted Timed Games to Emulate Memory

Weighted timed games are two-player zero-sum games played in a timedautomaton equipped with integer weights . In such weighted timed games, Minmay need finite memory to play (close to) optimally . In this work, we allow the players to use stochastic decisions, both in the choice of the timing oftransitions and of timing delays .…

Collaborative Multi Resource Allocation in Terrestrial Satellite Network Towards 6G

Terrestrial-satellite networks are envisioned to play a significant role inthe sixth-generation (6G) wireless networks . In such networks, hot air balloons are useful as they can relay the signals between satellites and groundstations . Most existing works assume the same height with the same minimum elevation angle to the satellites, which may not be practical due to possible route conflict with airplanes and otherflight equipment .…

Deep Extended Feedback Codes

A new deep-neural-network (DNN) based error correction encoder architecture called Deep Extended Feedback (DEF) is presented in this paper . The DEF architecture transmits an information message followed by a sequence of parity symbols which are generated based on the message as well as the observations of the past forward channel output to the transmitter through a feedback channel .…

Signal automata and hidden Markov models

A generic method for inferring a dynamical hidden Markov model from a timeseries is proposed . Under reasonable hypothesis, the model is updated inconstant time whenever a new measurement arrives .…

Streaming end to end speech recognition with jointly trained neural feature enhancement

We present a streaming end-to-end speech recognition model based on Monotonic Chunkwise Attention (MoCha) jointly trained with enhancement layers . We introduce two training strategies: GradualApplication of Enhanced Features (GAEF) and Gradual Reduction of Enhanced Loss (GREL) With GAEF, the model is initially trained using clean features .…

Multipath Graph Convolutional Neural Networks

Graph convolution networks have recently garnered a lot of attention forrepresentation learning on non-Euclidean feature spaces . In this work, we propose a novel Multipath Graphconvolutional neural network that aggregates the output of multiple different shallow networks . Results show that the proposed method attains increased test accuracy but also requires fewer trainingepochs to converge .…

Walk in the Cloud Learning Curves for Point Clouds Shape Analysis

Discrete point cloud objects lack sufficient shape descriptors of 3Dgeometries . In this paper, we present a novel method for aggregatinghypothetical curves in point clouds . Sequences of connected points (curves) areinitially grouped by taking guided walks in the point clouds, and then aggregated back to augment their point-wise features .…

Where and When Space Time Attention for Audio Visual Explanations

Recent advances in XAI provide explanations for models trained on still images . But when it comes to modeling multiplesensory modalities in a dynamic world, it remains underexplored how tomystify the mysterious dynamics of a complex multi-modal model . We propose a novel space-time attention network that uncovers the synergistic dynamics of audio and visual data overboth space and time .…

Towards Accountability in the Use of Artificial Intelligence for Public Administrations

We argue that distributed responsibility, inducedacceptance, and acceptance through ignorance constitute instances of imperfectdelegation when tasks are delegated to computationally-driven systems . We hold that both directpublic accountability via public transparency and indirect publicaccountability via transparency to auditors in public organizations can be bothinstrumentally ethically valuable and required as a matter of deontology from the principle of democratic self-government .…

MLP Mixer An all MLP Architecture for Vision

Convolutional Neural Networks (CNNs) are the go-to model for computer vision . Attention-based networks, such as the Vision Transformer, have also become popular . In this paper we show that while convolutions and attention are sufficient for good performance, neither of them are necessary .…

ZEN 2 0 Continue Training and Adaption for N gram Enhanced Text Encoders

Pre-trained text encoders have drawn sustaining attention in natural languageprocessing (NLP) They have shown their capability in obtaining promising results indifferent tasks . We propose topre-train n-gram-enhanced Encoders with a large volume of data and advanced techniques for training . We try to extend the encoder to different languages as well as different domains, where it is confirmed that the samarchitecture is applicable to these varying circumstances and new state-of-the-art performance is observed from a long list of NLP tasks across the languages and domains .…

Type IV II codes over Z4 constructed from generalized bent functions

A Type IV-II Z4-code is a self-dual code over Z4 with the property that allEuclidean weights are divisible by eight and all codewords have even Hammingweight . In this paper, we use generalized bent functions for a construction of self-orthogonal codes .…

We present an optimization-based approach to radiation treatment planning over time . Our approach formulates treatment planning as an optimal controlproblem with nonlinear patient health dynamics . As the formulation is nonconvex, we propose a method for obtaining an approximate solution .…

Structured Matrix Approximations via Tensor Decompositions

We provide a computational framework for approximating a class of structured matrices . Our approach has three steps: map the structured matrix totensors, use tensor compression algorithms, and map the compressed tensors back to obtain two different matrix representations . The resultingmatrix approximations are memory efficient, easy to compute with, and preserve error that is due to the compression in the Frobenius norm .…

Deterministic matrix sketches for low rank compression of high dimensional simulation data

Matrix sketching addresses this by reducing the input matrix to a smaller, but representative matrix via a low-dimensionallinear embedding . If the sketch matrix produced by the embedding capturessufficient geometric properties of the original matrix, then a near-optimal approximation may be obtained .…

NeuralLog a Neural Logic Language

The main goal of NeuralLog is to bridge logic programming and deep learning . The main advantages of neural networks are: to allow neural networks to be defined as logic programs; and to be able to handlenumeric attributes and functions .…

Small Sample Inferred Adaptive Recoding for Batched Network Coding

Batched network coding is a low-complexity network coding solution tofeedbackless multi-hop wireless packet network transmission with packet loss . Adaptiverecoding is a technique to adapt the fluctuation of packet loss by optimizing the number of recoded packets per batch to enhance the throughput .…

Combining Supervised and Un supervised Learning for Automatic Citrus Segmentation

Citrus segmentation is a key step of automatic citrus picking . Most current image segmentation approaches achieve good segmentation results bypixel-wise segmentation . These methods require alarge amount of annotated data, and do not consider continuous temporalchanges of citrus position in real-world applications .…

Effects of the COVID 19 Pandemic on Learning and Teaching a Case Study from Higher Education

In December 2019, the first case of SARS-CoV-2 infection was identified inWuhan, China . Since that day, COVID-19 has spread worldwide, affecting 153million people . Education has managed to adapt to therequirements and barriers implied by the impossibility to teach students face-to-face as it was done before .…

Leveraging Third Order Features in Skeleton Based Action Recognition

Recent skeleton-based actionrecognition methods extract features from 3D joint coordinates asspatial-temporal cues . We propose fusing third-orderfeatures in the form of angles into modern architectures, to robustly capture relationships between joints and body parts . This simple fusion with spatial-temporality graph neural networks achieves new state-of-the-artaccuracy in two large benchmarks, including NTU60 and NTU120, while employing fewer parameters and reduced run time .…

A Finer Calibration Analysis for Adversarial Robustness

We present a more general analysis of $H$-calibration for adversariallyrobust classification . By adopting a finer definition of calibration, we cancover settings beyond the restricted hypothesis sets studied in previous work . We both fix some previous calibration results (Bao et al.,…

Self Supervised Approach for Facial Movement Based Optical Flow

Deep learning-based optical flow techniques do not perform well for non-rigidmovements such as those found in faces . We hypothesize that learning opticalflow on face motion data will improve the quality of predicted flow on faces . The performance of FlowNetS trained on face images surpassed that of other opticalflow CNN architectures, demonstrating its usefulness.…

Speech Decomposition Based on a Hybrid Speech Model and Optimal Segmentation

In a hybrid speech model, both voiced and unvoiced components can coexist in a segment . We propose to adapt the segment length according to the signal local characteristics . A better extraction of the components is possible by taking into account the adaptive segmentation, compared to a fixed one .…

Regret Bounds for Stochastic Shortest Path Problems with Linear Function Approximation

We propose two algorithms for episodic stochastic shortest path problems withlinear function approximation . The first is computationally expensive butprovably obtains the same regret bound . The second is computatically efficient in practice . Both algorithms are based on an optimisticleast-squares version of value iteration analogous to the finite-horizonbackward induction approach from Jin et al.…

GT STORM Taming Sample Communication and Memory Complexities in Decentralized Non Convex Learning

Decentralized nonconvex optimization has received increasing attention in machine learning due to its advantages in system robustness, data privacy, and implementation simplicity . We show that to reach an$\epsilon^2$-stationary solution, the total number of sample evaluations of our algorithm is \$1/2 .…

TimeGym Debugging for Time Series Modeling in Python

TimeGym Forecasting Debugging Toolkit is a Python library fortesting and debugging time series forecasting pipelines . It provides generic tests for forecasting pipelines fresh out of the box . The library enables forecasters to apply aTest-Driven Development approach to forecast modeling, using specified oraclesto generate artificial data with noise .…

Comparison of Machine Learning Methods for Predicting Winter Wheat Yield in Germany

This study analyzed the performance of different machine learning methods for winter wheat yield prediction using extensive datasets of weather, soil, and phenology . The results indicated that nonlinear models such as deep neuralnetworks (DNN) and XGboost are more effective in finding the functionalrelationship between the crop yield and input data compared to linear models .…

BLM 17m A Large Scale Dataset for Black Lives Matter Topic Detection on Twitter

17 million tweets were collected from 25 May 2020 to21 August 2020 that covers 89 days from start of the George Floyd incident . The collected dataset is available at https://://://github.com/MeysamAsgariC/BLMT . We labeled thedataset by monitoring most trending news topics from global and local newspapers .…

Explaining how your AI system is fair

To implement fair machine learning in a sustainable way, choosing the rightfairness objective is key . The most appropriate fairness definition for an artificial intelligence system is a matter of ethical standards and legal requirements . In thisposition paper, we propose to use a decision tree as means to explain andjustify the implemented kind of fairness to the end users .…

Deep Neural Network for Musical Instrument Recognition using MFCCs

The task of efficient automatic music classification is of vital importance and forms the basis for various advanced applications of AI in the musical domain . Here we useuse only the mel-frequency cepstral coefficients (MFCCs) of the audio data . Our proposed model trains on the full London philharmonic orchestra dataset which contains twenty classes of instruments belonging to the four families viz.woodwinds,…

LACLICHEV Exploring the History of Climate Change in Latin America within Newspapers Digital Collections

LACLICHEV is based on data collections’ exploration techniques combined with informationretrieval, data analytics, and geographic querying and visualization . Thisenvironment provides tools for curating, exploring and analyzing historical articles, their description and location, and the vocabularies used to refer to meteorological events .…