PyTorchKR Papers

최신 논문

총 114편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2109.13104공식 코드 · Papers with Code

KNN-Averaging for Noisy Multi-objective Optimisation

Multi-objective optimisation is a popular approach for finding solutions to complex problems with large search spaces that reliably yields good optimisation results. However, with the rise of cyber-physical systems, emerges a new challenge of noisy fitness functions, whose objective value for a given configuration is non-deterministic, producing varying results on each execution. This leads to an optimisation process that is based on stochastically sampled information, ultimately favouring solutions with fitness values that have co-incidentally high outlier noise. In turn, the results are unfaithful due to their large discrepancies between sampled and expectable objective values. Motivated by our work on noisy automated driving systems, we present the results of our ongoing research to counteract the effect of noisy fitness functions without requiring repeated executions of each solution. Our method kNN-Avg identifies the k-nearest neighbours of a solution point and uses the weighted average value as a surrogate for its actually sampled fitness. We demonstrate the viability of kNN-Avg on common benchmark problems and show that it produces comparably good solutions whose fitness values are closer to the expected value.

arXiv:2109.08847공식 코드 · Papers with Code

Towards High-Quality Temporal Action Detection with Sparse Proposals

Temporal Action Detection (TAD) is an essential and challenging topic in video understanding, aiming to localize the temporal segments containing human action instances and predict the action categories. The previous works greatly rely upon dense candidates either by designing varying anchors or enumerating all the combinations of boundaries on video sequences; therefore, they are related to complicated pipelines and sensitive hand-crafted designs. Recently, with the resurgence of Transformer, query-based methods have tended to become the rising solutions for their simplicity and flexibility. However, there still exists a performance gap between query-based methods and well-established methods. In this paper, we identify the main challenge lies in the large variants of action duration and the ambiguous boundaries for short action instances; nevertheless, quadratic-computational global attention prevents query-based methods to build multi-scale feature maps. Towards high-quality temporal action detection, we introduce Sparse Proposals to interact with the hierarchical features. In our method, named SP-TAD, each proposal attends to a local segment feature in the temporal feature pyramid. The local interaction enables utilization of high-resolution features to preserve action instances details. Extensive experiments demonstrate the effectiveness of our method, especially under high tIoU thresholds. E.g., we achieve the state-of-the-art performance on THUMOS14 (45.7% on [email omitted], 33.4% on [email omitted] and 53.5% on mAP@Avg) and competitive results on ActivityNet-1.3 (32.99% on mAP@Avg). Code will be made available at https://github.com/wjn922/SP-TAD.

arXiv:2108.06069

Zero-shot Task Transfer for Invoice Extraction via Class-aware QA Ensemble

We present VESPA, an intentionally simple yet novel zero-shot system for layout, locale, and domain agnostic document extraction. In spite of the availability of large corpora of documents, the lack of labeled and validated datasets makes it a challenge to discriminatively train document extraction models for enterprises. We show that this problem can be addressed by simply transferring the information extraction (IE) task to a natural language Question-Answering (QA) task without engineering task-specific architectures. We demonstrate the effectiveness of our system by evaluating on a closed corpus of real-world retail and tax invoices with multiple complex layouts, domains, and geographies. The empirical evaluation shows that our system outperforms 4 prominent commercial invoice solutions that use discriminatively trained models with architectures specifically crafted for invoice extraction. We extracted 6 fields with zero upfront human annotation or training with an Avg. F1 of 87.50.

arXiv:2108.01469

Creation and Detection of German Voice Deepfakes

Synthesizing voice with the help of machine learning techniques has made rapid progress over the last years [1] and first high profile fraud cases have been recently reported [2]. Given the current increase in using conferencing tools for online teaching, we question just how easy (i.e. needed data, hardware, skill set) it would be to create a convincing voice fake. We analyse how much training data a participant (e.g. a student) would actually need to fake another participants voice (e.g. a professor). We provide an analysis of the existing state of the art in creating voice deep fakes, as well as offer detailed technical guidance and evidence of just how much effort is needed to copy a voice. A user study with more than 100 participants shows how difficult it is to identify real and fake voice (on avg. only 37 percent can distinguish between real and fake voice of a professor). With a focus on German language and an online teaching environment we discuss the societal implications as well as demonstrate how to use machine learning techniques to possibly detect such fakes.

arXiv:2106.08564

Adaptive Visibility Graph Neural Network and It's Application in Modulation Classification

Our digital world is full of time series and graphs which capture the various aspects of many complex systems. Traditionally, there are respective methods in processing these two different types of data, e.g., Recurrent Neural Network (RNN) and Graph Neural Network (GNN), while in recent years, time series could be mapped to graphs by using the techniques such as Visibility Graph (VG), so that researchers can use graph algorithms to mine the knowledge in time series. Such mapping methods establish a bridge between time series and graphs, and have high potential to facilitate the analysis of various real-world time series. However, the VG method and its variants are just based on fixed rules and thus lack of flexibility, largely limiting their application in reality. In this paper, we propose an Adaptive Visibility Graph (AVG) algorithm that can adaptively map time series into graphs, based on which we further establish an end-to-end classification framework AVGNet, by utilizing GNN model DiffPool as the classifier. We then adopt AVGNet for radio signal modulation classification which is an important task in the field of wireless communication. The simulations validate that AVGNet outperforms a series of advanced deep learning methods, achieving the state-of-the-art performance in this task.

arXiv:2105.14538

Longer Version for "Deep Context-Encoding Network for Retinal Image Captioning"

Automatically generating medical reports for retinal images is one of the promising ways to help ophthalmologists reduce their workload and improve work efficiency. In this work, we propose a new context-driven encoding network to automatically generate medical reports for retinal images. The proposed model is mainly composed of a multi-modal input encoder and a fused-feature decoder. Our experimental results show that our proposed method is capable of effectively leveraging the interactive information between the input image and context, i.e., keywords in our case. The proposed method creates more accurate and meaningful reports for retinal images than baseline models and achieves state-of-the-art performance. This performance is shown in several commonly used metrics for the medical report generation task: BLEU-avg (+16%), CIDEr (+10.2%), and ROUGE (+8.6%).

arXiv:2105.13965

Revitalizing Optimization for 3D Human Pose and Shape Estimation: A Sparse Constrained Formulation

We propose a novel sparse constrained formulation and from it derive a real-time optimization method for 3D human pose and shape estimation. Our optimization method, SCOPE (Sparse Constrained Optimization for 3D human Pose and shapE estimation), is orders of magnitude faster (avg. 4 ms convergence) than existing optimization methods, while being mathematically equivalent to their dense unconstrained formulation under mild assumptions. We achieve this by exploiting the underlying sparsity and constraints of our formulation to efficiently compute the Gauss-Newton direction. We show that this computation scales linearly with the number of joints and measurements of a complex 3D human model, in contrast to prior work where it scales cubically due to their dense unconstrained formulation. Based on our optimization method, we present a real-time motion capture framework that estimates 3D human poses and shapes from a single image at over 30 FPS. In benchmarks against state-of-the-art methods on multiple public datasets, our framework outperforms other optimization methods and achieves competitive accuracy against regression methods. Project page with code and videos: https://sites.google.com/view/scope-human/.

arXiv:2104.12471

Contextualized Keyword Representations for Multi-modal Retinal Image Captioning

Medical image captioning automatically generates a medical description to describe the content of a given medical image. A traditional medical image captioning model creates a medical description only based on a single medical image input. Hence, an abstract medical description or concept is hard to be generated based on the traditional approach. Such a method limits the effectiveness of medical image captioning. Multi-modal medical image captioning is one of the approaches utilized to address this problem. In multi-modal medical image captioning, textual input, e.g., expert-defined keywords, is considered as one of the main drivers of medical description generation. Thus, encoding the textual input and the medical image effectively are both important for the task of multi-modal medical image captioning. In this work, a new end-to-end deep multi-modal medical image captioning model is proposed. Contextualized keyword representations, textual feature reinforcement, and masked self-attention are used to develop the proposed approach. Based on the evaluation of the existing multi-modal medical image captioning dataset, experimental results show that the proposed model is effective with the increase of +53.2% in BLEU-avg and +18.6% in CIDEr, compared with the state-of-the-art method.

arXiv:2103.09512공식 코드 · Papers with Code

CNN Model & Tuning for Global Road Damage Detection

This paper provides a report on our solution including model selection, tuning strategy and results obtained for Global Road Damage Detection Challenge. This Big Data Cup Challenge was held as a part of IEEE International Conference on Big Data 2020. We assess single and multi-stage network architectures for object detection and provide a benchmark using popular state-of-the-art open-source PyTorch frameworks like Detectron2 and Yolov5. Data preparation for provided Road Damage training dataset, captured using smartphone camera from Czech, India and Japan is discussed. We studied the effect of training on a per country basis with respect to a single generalizable model. We briefly describe the tuning strategy for the experiments conducted on two-stage Faster R-CNN with Deep Residual Network (Resnet) and Feature Pyramid Network (FPN) backbone. Additionally, we compare this to a one-stage Yolov5 model with Cross Stage Partial Network (CSPNet) backbone. We show a mean F1 score of 0.542 on Test2 and 0.536 on Test1 datasets using a multi-stage Faster R-CNN model, with Resnet-50 and Resnet-101 backbones respectively. This shows the generalizability of the Resnet-50 model when compared to its more complex counterparts. Experiments were conducted using Google Colab having K80 and a Linux PC with 1080Ti, NVIDIA consumer grade GPU. A PyTorch based Detectron2 code to preprocess, train, test and submit the Avg F1 score to is made available at https://github.com/vishwakarmarhl/rdd2020

arXiv:2103.09019

Intelligent colocation of HPC workloads

Many HPC applications suffer from a bottleneck in the shared caches, instruction execution units, I/O or memory bandwidth, even though the remaining resources may be underutilized. It is hard for developers and runtime systems to ensure that all critical resources are fully exploited by a single application, so an attractive technique for increasing HPC system utilization is to colocate multiple applications on the same server. When applications share critical resources, however, contention on shared resources may lead to reduced application performance. In this paper, we show that server efficiency can be improved by first modeling the expected performance degradation of colocated applications based on measured hardware performance counters, and then exploiting the model to determine an optimized mix of colocated applications. This paper presents a new intelligent resource manager and makes the following contributions: (1) a new machine learning model to predict the performance degradation of colocated applications based on hardware counters and (2) an intelligent scheduling scheme deployed on an existing resource manager to enable application co-scheduling with minimum performance degradation. Our results show that our approach achieves performance improvements of 7% (avg) and 12% (max) compared to the standard policy commonly used by existing job managers.

arXiv:2103.00424

SpikeDyn: A Framework for Energy-Efficient Spiking Neural Networks with Continual and Unsupervised Learning Capabilities in Dynamic Environments

Spiking Neural Networks (SNNs) bear the potential of efficient unsupervised and continual learning capabilities because of their biological plausibility, but their complexity still poses a serious research challenge to enable their energy-efficient design for resource-constrained scenarios (like embedded systems, IoT-Edge, etc.). We propose SpikeDyn, a comprehensive framework for energy-efficient SNNs with continual and unsupervised learning capabilities in dynamic environments, for both the training and inference phases. It is achieved through the following multiple diverse mechanisms: 1) reduction of neuronal operations, by replacing the inhibitory neurons with direct lateral inhibitions; 2) a memory- and energy-constrained SNN model search algorithm that employs analytical models to estimate the memory footprint and energy consumption of different candidate SNN models and selects a Pareto-optimal SNN model; and 3) a lightweight continual and unsupervised learning algorithm that employs adaptive learning rates, adaptive membrane threshold potential, weight decay, and reduction of spurious updates. Our experimental results show that, for a network with 400 excitatory neurons, our SpikeDyn reduces the energy consumption on average by 51% for training and by 37% for inference, as compared to the state-of-the-art. Due to the improved learning algorithm, SpikeDyn provides on avg. 21% accuracy improvement over the state-of-the-art, for classifying the most recently learned task, and by 8% on average for the previously learned tasks.

arXiv:2103.00384

Adaptive Regularized Submodular Maximization

In this paper, we study the problem of maximizing the difference between an adaptive submodular (revenue) function and an non-negative modular (cost) function under the adaptive setting. The input of our problem is a set of $n$ items, where each item has a particular state drawn from some known prior distribution $p$. The revenue function $g$ is defined over items and states, and the cost function $c$ is defined over items, i.e., each item has a fixed cost. The state of each item is unknown initially, one must select an item in order to observe its realized state. A policy $\pi$ specifies which item to pick next based on the observations made so far. Denote by $g_{avg}(\pi)$ the expected revenue of $\pi$ and let $c_{avg}(\pi)$ denote the expected cost of $\pi$. Our objective is to identify the best policy $\pi^o\in \arg\max_{\pi}g_{avg}(\pi)-c_{avg}(\pi)$ under a $k$-cardinality constraint. Since our objective function can take on both negative and positive values, the existing results of submodular maximization may not be applicable. To overcome this challenge, we develop a series of effective solutions with performance grantees. Let $\pi^o$ denote the optimal policy. For the case when $g$ is adaptive monotone and adaptive submodular, we develop an effective policy $\pi^l$ such that $g_{avg}(\pi^l) - c_{avg}(\pi^l) \geq (1-\frac{1}{e}-\epsilon)g_{avg}(\pi^o) - c_{avg}(\pi^o)$, using only $O(n\epsilon^{-2}\log \epsilon^{-1})$ value oracle queries. For the case when $g$ is adaptive submodular, we present a randomized policy $\pi^r$ such that $g_{avg}(\pi^r) - c_{avg}(\pi^r) \geq \frac{1}{e}g_{avg}(\pi^o) - c_{avg}(\pi^o)$.

arXiv:2012.02387

A Variant of Gradient Descent Algorithm Based on Gradient Averaging

In this work, we study an optimizer, Grad-Avg to optimize error functions. We establish the convergence of the sequence of iterates of Grad-Avg mathematically to a minimizer (under boundedness assumption). We apply Grad-Avg along with some of the popular optimizers on regression as well as classification tasks. In regression tasks, it is observed that the behaviour of Grad-Avg is almost identical with Stochastic Gradient Descent (SGD). We present a mathematical justification of this fact. In case of classification tasks, it is observed that the performance of Grad-Avg can be enhanced by suitably scaling the parameters. Experimental results demonstrate that Grad-Avg converges faster than the other state-of-the-art optimizers for the classification task on two benchmark datasets.

arXiv:1908.00257

Quantifying horizon dependence of asset prices: a cluster entropy approach

Market dynamic is quantified in terms of the entropy $S(\tau,n)$ of the clusters formed by the intersections between the series of the prices $p_t$ and the moving average $\widetilde{p}_{t,n}$. The entropy $S(\tau,n)$ is defined according to Shannon as $\sum P(\tau,n)\log P(\tau,n),$ with $P(\tau,n)$ the probability for the cluster to occur with duration $\tau$. \par The investigation is performed on high-frequency data of the Nasdaq Composite, Dow Jones Industrial Avg and Standard \& Poor 500 indexes downloaded from the Bloomberg terminal. The cluster entropy $S(\tau,n)$ is analysed in raw and sampled data over a broad range of temporal horizons $M$ varying from one to twelve months over the year 2018. The cluster entropy $S(\tau,n)$ is integrated over the cluster duration $\tau$ to yield the Market Dynamic Index $I(M,n)$, a synthetic figure of price dynamics. A systematic dependence of the cluster entropy $S(\tau,n)$ and the Market Dynamic Index $I(M,n)$ on the temporal horizon $M$ is evidenced. \par Finally, the Market Horizon Dependence}, defined as $H(M,n)=I(M,n)-I(1,n)$, is compared with the horizon dependence of the pricing kernel with different representative agents obtained via a Kullback-Leibler entropy approach. The Market Horizon Dependence $H(M,n)$ of the three assets is compared against the values obtained by implementing the cluster entropy $S(\tau,n)$ approach on artificially generated series (Fractional Brownian Motion).

arXiv:2011.01181

QMUL-SDS @ SardiStance: Leveraging Network Interactions to Boost Performance on Stance Detection using Knowledge Graphs

This paper presents our submission to the SardiStance 2020 shared task, describing the architecture used for Task A and Task B. While our submission for Task A did not exceed the baseline, retraining our model using all the training tweets, showed promising results leading to (f-avg 0.601) using bidirectional LSTM with BERT multilingual embedding for Task A. For our submission for Task B, we ranked 6th (f-avg 0.709). With further investigation, our best experimented settings increased performance from (f-avg 0.573) to (f-avg 0.733) with same architecture and parameter settings and after only incorporating social interaction features -- highlighting the impact of social interaction on the model's performance.

arXiv:2003.09505

A Reliability-aware Multi-armed Bandit Approach to Learn and Select Users in Demand Response

One challenge in the optimization and control of societal systems is to handle the unknown and uncertain user behavior. This paper focuses on residential demand response (DR) and proposes a closed-loop learning scheme to address these issues. In particular, we consider DR programs where an aggregator calls upon residential users to change their demand so that the total load adjustment is close to a target value. To learn and select the right users, we formulate the DR problem as a combinatorial multi-armed bandit (CMAB) problem with a reliability objective. We propose a learning algorithm: CUCB-Avg (Combinatorial Upper Confidence Bound-Average), which utilizes both upper confidence bounds and sample averages to balance the tradeoff between exploration (learning) and exploitation (selecting). We consider both a fixed time-invariant target and time-varying targets, and show that CUCB-Avg achieves $O(\log T)$ and $O(\sqrt{T \log(T)})$ regrets respectively. Finally, we numerically test our algorithms using synthetic and real data, and demonstrate that our CUCB-Avg performs significantly better than the classic CUCB and also better than Thompson Sampling.

arXiv:2009.12013공식 코드 · Papers with Code

Revealing the Myth of Higher-Order Inference in Coreference Resolution

This paper analyzes the impact of higher-order inference (HOI) on the task of coreference resolution. HOI has been adapted by almost all recent coreference resolution models without taking much investigation on its true effectiveness over representation learning. To make a comprehensive analysis, we implement an end-to-end coreference system as well as four HOI approaches, attended antecedent, entity equalization, span clustering, and cluster merging, where the latter two are our original methods. We find that given a high-performing encoder such as SpanBERT, the impact of HOI is negative to marginal, providing a new perspective of HOI to this task. Our best model using cluster merging shows the Avg-F1 of 80.2 on the CoNLL 2012 shared task dataset in English.

arXiv:2008.08474

Human Body Model Fitting by Learned Gradient Descent

We propose a novel algorithm for the fitting of 3D human shape to images. Combining the accuracy and refinement capabilities of iterative gradient-based optimization techniques with the robustness of deep neural networks, we propose a gradient descent algorithm that leverages a neural network to predict the parameter update rule for each iteration. This per-parameter and state-aware update guides the optimizer towards a good solution in very few steps, converging in typically few steps. During training our approach only requires MoCap data of human poses, parametrized via SMPL. From this data the network learns a subspace of valid poses and shapes in which optimization is performed much more efficiently. The approach does not require any hard to acquire image-to-3D correspondences. At test time we only optimize the 2D joint re-projection error without the need for any further priors or regularization terms. We show empirically that this algorithm is fast (avg. 120ms convergence), robust to initialization and dataset, and achieves state-of-the-art results on public evaluation datasets including the challenging 3DPW in-the-wild benchmark (improvement over SMPLify 45%) and also approaches using image-to-3D correspondences

arXiv:2006.11416공식 코드 · Papers with Code

A Symbolic Temporal Pooling method for Video-based Person Re-Identification

In video-based person re-identification, both the spatial and temporal features are known to provide orthogonal cues to effective representations. Such representations are currently typically obtained by aggregating the frame-level features using max/avg pooling, at different points of the models. However, such operations also decrease the amount of discriminating information available, which is particularly hazardous in case of poor separability between the different classes. To alleviate this problem, this paper introduces a symbolic temporal pooling method, where frame-level features are represented in the distribution valued symbolic form, yielding from fitting an Empirical Cumulative Distribution Function (ECDF) to each feature. Also, considering that the original triplet loss formulation cannot be applied directly to this kind of representations, we introduce a symbolic triplet loss function that infers the similarity between two symbolic objects. Having carried out an extensive empirical evaluation of the proposed solution against the state-of-the-art, in four well known data sets (MARS, iLIDS-VID, PRID2011 and P-DESTRE), the observed results point for consistent improvements in performance over the previous best performing techniques.

arXiv:2005.08367공식 코드 · Papers with Code

DEXA: Supporting Non-Expert Annotators with Dynamic Examples from Experts

The success of crowdsourcing based annotation of text corpora depends on ensuring that crowdworkers are sufficiently well-trained to perform the annotation task accurately. To that end, a frequent approach to train annotators is to provide instructions and a few example cases that demonstrate how the task should be performed (referred to as the CONTROL approach). These globally defined "task-level examples", however, (i) often only cover the common cases that are encountered during an annotation task; and (ii) require effort from crowdworkers during the annotation process to find the most relevant example for the currently annotated sample. To overcome these limitations, we propose to support workers in addition to task-level examples, also with "task-instance level" examples that are semantically similar to the currently annotated data sample (referred to as Dynamic Examples for Annotation, DEXA). Such dynamic examples can be retrieved from collections previously labeled by experts, which are usually available as gold standard dataset. We evaluate DEXA on a complex task of annotating participants, interventions, and outcomes (known as PIO) in sentences of medical studies. The dynamic examples are retrieved using BioSent2Vec, an unsupervised semantic sentence similarity method specific to the biomedical domain. Results show that (i) workers of the DEXA approach reach on average much higher agreements (Cohen's Kappa) to experts than workers of the the CONTROL approach (avg. of 0.68 to experts in DEXA vs. 0.40 in CONTROL); (ii) already three per majority voting aggregated annotations of the DEXA approach reach substantial agreements to experts of 0.78/0.75/0.69 for P/I/O (in CONTROL 0.73/0.58/0.46). Finally, (iii) we acquire explicit feedback from workers and show that in the majority of cases (avg. 72%) workers find the dynamic examples useful.