Incorporating artificial intelligence and machine learning (AI/ML) methods within the 5G wireless standard promises autonomous network behavior and ultra-low-latency reconfiguration. However, the effort so far has purely focused on learning from radio frequency (RF) signals. Future standards and next-generation (nextG) networks beyond 5G will have two significant evolutions over the state-of-the-art 5G implementations: (i) massive number of antenna elements, scaling up to hundreds-to-thousands in number, and (ii) inclusion of AI/ML in the critical path of the network reconfiguration process that can access sensor feeds from a variety of RF and non-RF sources. While the former allows unprecedented flexibility in 'beamforming', where signals combine constructively at a target receiver, the latter enables the network with enhanced situation awareness not captured by a single and isolated data modality. This survey presents a thorough analysis of the different approaches used for beamforming today, focusing on mmWave bands, and then proceeds to make a compelling case for considering non-RF sensor data from multiple modalities, such as LiDAR, Radar, GPS for increasing beamforming directional accuracy and reducing processing time. This so called idea of multimodal beamforming will require deep learning based fusion techniques, which will serve to augment the current RF-only and classical signal processing methods that do not scale well for massive antenna arrays. The survey describes relevant deep learning architectures for multimodal beamforming, identifies computational challenges and the role of edge computing in this process, dataset generation tools, and finally, lists open challenges that the community should tackle to realize this transformative vision of the future of beamforming.
Robotic cutting of soft materials is critical for applications such as food processing, household automation, and surgical manipulation. As in other areas of robotics, simulators can facilitate controller verification, policy learning, and dataset generation. Moreover, differentiable simulators can enable gradient-based optimization, which is invaluable for calibrating simulation parameters and optimizing controllers. In this work, we present DiSECt: the first differentiable simulator for cutting soft materials. The simulator augments the finite element method (FEM) with a continuous contact model based on signed distance fields (SDF), as well as a continuous damage model that inserts springs on opposite sides of the cutting plane and allows them to weaken until zero stiffness, enabling crack formation. Through various experiments, we evaluate the performance of the simulator. We first show that the simulator can be calibrated to match resultant forces and deformation fields from a state-of-the-art commercial solver and real-world cutting datasets, with generality across cutting velocities and object instances. We then show that Bayesian inference can be performed efficiently by leveraging the differentiability of the simulator, estimating posteriors over hundreds of parameters in a fraction of the time of derivative-free methods. Next, we illustrate that control parameters in the simulation can be optimized to minimize cutting forces via lateral slicing motions. Finally, we conduct experiments on a real robot arm equipped with a slicing knife to infer simulation parameters from force measurements. By optimizing the slicing motion of the knife, we show on fruit cutting scenarios that the average knife force can be reduced by more than 40% compared to a vertical cutting motion. We publish code and additional materials on our project website at https://diff-cutting-sim.github.io.
Deep decarbonization of the energy sector will require massive penetration of stochastic renewable energy resources and an enormous amount of grid asset coordination; this represents a challenging paradigm for the power system operators who are tasked with maintaining grid stability and security in the face of such changes. With its ability to learn from complex datasets and provide predictive solutions on fast timescales, machine learning (ML) is well-posed to help overcome these challenges as power systems transform in the coming decades. In this work, we outline five key challenges (dataset generation, data pre-processing, model training, model assessment, and model embedding) associated with building trustworthy ML models which learn from physics-based simulation data. We then demonstrate how linking together individual modules, each of which overcomes a respective challenge, at sequential stages in the machine learning pipeline can help enhance the overall performance of the training process. In particular, we implement methods that connect different elements of the learning pipeline through feedback, thus "closing the loop" between model training, performance assessments, and re-training. We demonstrate the effectiveness of this framework, its constituent modules, and its feedback connections by learning the N-1 small-signal stability margin associated with a detailed model of a proposed North Sea Wind Power Hub system.
Accurate recommendation and reliable explanation are two key issues for modern recommender systems. However, most recommendation benchmarks only concern the prediction of user-item ratings while omitting the underlying causes behind the ratings. For example, the widely-used Yahoo!R3 dataset contains little information on the causes of the user-movie ratings. A solution could be to conduct surveys and require the users to provide such information. In practice, the user surveys can hardly avoid compliance issues and sparse user responses, which greatly hinders the exploration of causality-based recommendation. To better support the studies of causal inference and further explanations in recommender systems, we propose a novel semi-synthetic data generation framework for recommender systems where causal graphical models with missingness are employed to describe the causal mechanism of practical recommendation scenarios. To illustrate the use of our framework, we construct a semi-synthetic dataset with Causal Tags And Ratings (CTAR), based on the movies as well as their descriptive tags and rating information collected from a famous movie rating website. Using the collected data and the causal graph, the user-item-ratings and their corresponding user-item-tags are automatically generated, which provides the reasons (selected tags) why the user rates the items. Descriptive statistics and baseline results regarding the CTAR dataset are also reported. The proposed data generation framework is not limited to recommendation, and the released APIs can be used to generate customized datasets for other research tasks.
There is a growing interest in dataset generation recently due to the superior generative capacity of large pre-trained language models (PLMs). In this paper, we study a flexible and efficient zero-short learning method, \textsc{ZeroGen}. Given a zero-shot task, we first generate a dataset from scratch using PLMs in an unsupervised manner. Then, we train a tiny task model (e.g., LSTM) under the supervision of the synthesized dataset. This approach allows highly efficient inference as the final task model only has orders of magnitude fewer parameters comparing to PLMs (e.g., GPT2-XL). Apart from being annotation-free and efficient, we argue that \textsc{ZeroGen} can also provide useful insights from the perspective of data-free model-agnostic knowledge distillation, and unreferenced text generation evaluation. Experiments and analysis on different NLP tasks, namely, text classification, question answering, and natural language inference, show the effectiveness of \textsc{ZeroGen}.
In recent years, deep learning techniques have been introduced into the field of trajectory optimization to improve convergence and speed. Training such models requires large trajectory datasets. However, the convergence of low thrust (LT) optimizations is unpredictable before the optimization process ends. For randomly initialized low thrust transfer data generation, most of the computation power will be wasted on optimizing infeasible low thrust transfers, which leads to an inefficient data generation process. This work proposes a deep neural network (DNN) classifier to accurately identify feasible LT transfer prior to the optimization process. The DNN-classifier achieves an overall accuracy of 97.9%, which has the best performance among the tested algorithms. The accurate low-thrust trajectory feasibility identification can avoid optimization on undesired samples, so that the majority of the optimized samples are LT trajectories that converge. This technique enables efficient dataset generation for different mission scenarios with different spacecraft configurations.
Recent advancements in computer vision have seen a rise in the prominence of applications using neural networks to understand human poses. However, while accuracy has been steadily increasing on State-of-the-Art datasets, these datasets often do not address the challenges seen in real-world applications. These challenges are dealing with people distant from the camera, people in crowds, and heavily occluded people. As a result, many real-world applications have trained on data that does not reflect the data present in deployment, leading to significant underperformance. This article presents ADG-Pose, a method for automatically generating datasets for real-world human pose estimation. These datasets can be customized to determine person distances, crowdedness, and occlusion distributions. Models trained with our method are able to perform in the presence of these challenges where those trained on other datasets fail. Using ADG-Pose, end-to-end accuracy for real-world skeleton-based action recognition sees a 20% increase on scenes with moderate distance and occlusion levels, and a 4X increase on distant scenes where other models failed to perform better than random.
Synthetic data and simulators have the potential to markedly improve the performance and robustness of recommendation systems. These approaches have already had a beneficial impact in other machine-learning driven fields. We identify and discuss a key trade-off between data fidelity and privacy in the past work on synthetic data and simulators for recommendation systems. For the important use case of predicting algorithm rankings on real data from synthetic data, we provide motivation and current successes versus limitations. Finally we outline a number of exciting future directions for recommendation systems that we believe deserve further attention and work, including mixing real and synthetic data, feedback in dataset generation, robust simulations, and privacy-preserving methods.
While the English virtual assistants have achieved exciting performance with an enormous amount of training resources, the needs of non-English-speakers have not been satisfied well. Up to Dec 2021, Alexa, one of the most popular smart speakers around the world, is able to support 9 different languages [1], while there are thousands of languages in the world, 91 of which are spoken by more than 10 million people according to statistics published in 2019 [2]. However, training a virtual assistant in other languages than English is often more difficult, especially for those low-resource languages. The lack of high-quality training data restricts the performance of models, resulting in poor user satisfaction. Therefore, we devise an efficient and effective training solution for multilingual task-orientated dialogue systems, using the same dataset generation pipeline and end-to-end dialogue system architecture as BiToD[5], which adopted some key design choices for a minimalistic natural language design where formal dialogue states are used in place of natural language inputs. This reduces the room for error brought by weaker natural language models, and ensures the model can correctly extract the essential slot values needed to perform dialogue state tracking (DST). Our goal is to reduce the amount of natural language encoded at each turn, and the key parameter we investigate is the number of turns (H) to feed as history to model. We first explore the turning point where increasing H begins to yield limiting returns on the overall performance. Then we examine whether the examples a model with small H gets wrong can be categorized in a way for the model to do few-shot finetuning on. Lastly, will explore the limitations of this approach, and whether there is a certain type of examples that this approach will not be able to resolve.
Optical wireless communication offers unprecedented communication speeds that can support the massive use of the Internet on a daily basis. In indoor environments, optical wireless networks are usually multi-user multiple-input multiple-output (MU-MIMO) systems, where a high number of optical access points (APs) is required to ensure coverage. In this work, a laser-based optical wireless network is considered for serving multiple users. Moreover, blind inference alignment (BIA) is implemented to achieve a high degree of freedom (DoF) without the need for channel state information (CSI) at transmitters, which is difficult to provide in such wireless networks. Then, an objective function is defined to allocate the resources of the network taking into consideration the requirements of users and the available resources. This optimization problem can be solved through exhaustive search or distributed algorithms. However, a practical algorithm that provides immediate solutions in real time scenarios is required. In this context, an artificial neural network (ANN) model is derived in order to obtain a sub-optimal solution with low computational time. The implementation of the ANN model involves three important steps, dataset generation, offline training, and real time application. The results show that the trained ANN model provides a significant solution close to the optimal one.
Building of data for quality estimation (QE) training is expensive and requires significant human labor. In this study, we focus on a data-centric approach while performing QE, and subsequently propose a fully automatic pseudo-QE dataset generation tool that generates QE datasets by receiving only monolingual or parallel corpus as the input. Consequently, the QE performance is enhanced either by data augmentation or by encouraging multiple language pairs to exploit the applicability of QE. Further, we intend to publicly release this user friendly QE dataset generation tool as we believe this tool provides a new, inexpensive method to the community for developing QE datasets.
Digital image correlation (DIC) has become an industry standard to retrieve accurate displacement and strain measurement in tensile testing and other material characterization. Though traditional DIC offers a high precision estimation of deformation for general tensile testing cases, the prediction becomes unstable at large deformation or when the speckle patterns start to tear. In addition, traditional DIC requires a long computation time and often produces a low spatial resolution output affected by filtering and speckle pattern quality. To address these challenges, we propose a new deep learning-based DIC approach--Deep DIC, in which two convolutional neural networks, DisplacementNet and StrainNet, are designed to work together for end-to-end prediction of displacements and strains. DisplacementNet predicts the displacement field and adaptively tracks a region of interest. StrainNet predicts the strain field directly from the image input without relying on the displacement prediction, which significantly improves the strain prediction accuracy. A new dataset generation method is developed to synthesize a realistic and comprehensive dataset, including the generation of speckle patterns and the deformation of the speckle image with synthetic displacement fields. Though trained on synthetic datasets only, Deep DIC gives highly consistent and comparable predictions of displacement and strain with those obtained from commercial DIC software for real experiments, while it outperforms commercial software with very robust strain prediction even at large and localized deformation and varied pattern qualities. In addition, Deep DIC is capable of real-time prediction of deformation with a calculation time down to milliseconds.
Vaccine hesitancy, which has recently been driven by online narratives, significantly degrades the efficacy of vaccination strategies, such as those for COVID-19. Despite broad agreement in the medical community about the safety and efficacy of available vaccines, a large number of social media users continue to be inundated with false information about vaccines and are indecisive or unwilling to be vaccinated. The goal of this study is to better understand anti-vaccine sentiment by developing a system capable of automatically identifying the users responsible for spreading anti-vaccine narratives. We introduce a publicly available Python package capable of analyzing Twitter profiles to assess how likely that profile is to share anti-vaccine sentiment in the future. The software package is built using text embedding methods, neural networks, and automated dataset generation and is trained on several million tweets. We find this model can accurately detect anti-vaccine users up to a year before they tweet anti-vaccine hashtags or keywords. We also show examples of how text analysis helps us understand anti-vaccine discussions by detecting moral and emotional differences between anti-vaccine spreaders on Twitter and regular users. Our results will help researchers and policy-makers understand how users become anti-vaccine and what they discuss on Twitter. Policy-makers can utilize this information for better targeted campaigns that debunk harmful anti-vaccination myths.
This paper presents a method for automatic generation of a training dataset for a deep convolutional neural network used for playing card detection. The solution allows to skip the time-consuming processes of manual image collecting and labelling recognised objects. The YOLOv4 network trained on the generated dataset achieved an efficiency of 99.8% in the cards detection task. The proposed method is a part of a project that aims to automate the process of broadcasting duplicate bridge competitions using a vision system and neural networks.
Researchers often spend weeks sifting through decades of unlabeled satellite imagery(on NASA Worldview) in order to develop datasets on which they can start conducting research. We developed an interactive, scalable and fast image similarity search engine (which can take one or more images as the query image) that automatically sifts through the unlabeled dataset reducing dataset generation time from weeks to minutes. In this work, we describe key components of the end to end pipeline. Our similarity search system was created to be able to identify similar images from a potentially petabyte scale database that are similar to an input image, and for this we had to break down each query image into its features, which were generated by a classification layer stripped CNN trained in a supervised manner. To store and search these features efficiently, we had to make several scalability improvements. To improve the speed, reduce the storage, and shrink memory requirements for embedding search, we add a fully connected layer to our CNN make all images into a 128 length vector before entering the classification layers. This helped us compress the size of our image features from 2048 (for ResNet, which was initially tried as our featurizer) to 128 for our new custom model. Additionally, we utilize existing approximate nearest neighbor search libraries to significantly speed up embedding search. Our system currently searches over our entire database of images at 5 seconds per query on a single virtual machine in the cloud. In the future, we would like to incorporate a SimCLR based featurizing model which could be trained without any labelling by a human (since the classification aspect of the model is irrelevant to this use case).
Realizing edge intelligence consists of sensing, communication, training, and inference stages. Conventionally, the sensing and communication stages are executed sequentially, which results in excessive amount of dataset generation and uploading time. This paper proposes to accelerate edge intelligence via integrated sensing and communication (ISAC). As such, the sensing and communication stages are merged so as to make the best use of the wireless signals for the dual purpose of dataset generation and uploading. However, ISAC also introduces additional interference between sensing and communication functionalities. To address this challenge, this paper proposes a classification error minimization formulation to design the ISAC beamforming and time allocation. The globally optimal solution is derived via the rank-1 guaranteed semidefinite relaxation, and performance analysis is performed to quantify the ISAC gain over that of conventional edge intelligence. Simulation results are provided to verify the effectiveness of the proposed ISAC-assisted edge intelligence system. Interestingly, we find that ISAC is always beneficial, when the duration of generating a sample is more than the duration of uploading a sample. Otherwise, the ISAC gain can vanish or even be negative. Nevertheless, we still derive a sufficient condition, under which a positive ISAC gain is feasible.
We present SynPick, a synthetic dataset for dynamic scene understanding in bin-picking scenarios. In contrast to existing datasets, our dataset is both situated in a realistic industrial application domain -- inspired by the well-known Amazon Robotics Challenge (ARC) -- and features dynamic scenes with authentic picking actions as chosen by our picking heuristic developed for the ARC 2017. The dataset is compatible with the popular BOP dataset format. We describe the dataset generation process in detail, including object arrangement generation and manipulation simulation using the NVIDIA PhysX physics engine. To cover a large action space, we perform untargeted and targeted picking actions, as well as random moving actions. To establish a baseline for object perception, a state-of-the-art pose estimation approach is evaluated on the dataset. We demonstrate the usefulness of tracking poses during manipulation instead of single-shot estimation even with a naive filtering approach. The generator source code and dataset are publicly available.
This paper introduces a shoebox room simulator able to systematically generate synthetic datasets of binaural room impulse responses (BRIRs) given an arbitrary set of head-related transfer functions (HRTFs). The evaluation of machine hearing algorithms frequently requires BRIR datasets in order to simulate the acoustics of any environment. However, currently available solutions typically consider only HRTFs measured on dummy heads, which poorly characterize the high variability in spatial sound perception. Our solution allows to integrate a room impulse response (RIR) simulator with different HRTF sets represented in Spatially Oriented Format for Acoustics (SOFA). The source code and the compiled binaries for different operating systems allow to both advanced and non-expert users to benefit from our toolbox, see https://github.com/spatialaudiotools/sofamyroom/ .
If two sentences have the same meaning, it should follow that they are equivalent in their inferential properties, i.e., each sentence should textually entail the other. However, many paraphrase datasets currently in widespread use rely on a sense of paraphrase based on word overlap and syntax. Can we teach them instead to identify paraphrases in a way that draws on the inferential properties of the sentences, and is not over-reliant on lexical and syntactic similarities of a sentence pair? We apply the adversarial paradigm to this question, and introduce a new adversarial method of dataset creation for paraphrase identification: the Adversarial Paraphrasing Task (APT), which asks participants to generate semantically equivalent (in the sense of mutually implicative) but lexically and syntactically disparate paraphrases. These sentence pairs can then be used both to test paraphrase identification models (which get barely random accuracy) and then improve their performance. To accelerate dataset generation, we explore automation of APT using T5, and show that the resulting dataset also improves accuracy. We discuss implications for paraphrase detection and release our dataset in the hope of making paraphrase detection models better able to detect sentence-level meaning equivalence.
Human life is populated with articulated objects. Current Category-level Articulation Pose Estimation (CAPE) methods are studied under the single-instance setting with a fixed kinematic structure for each category. Considering these limitations, we reform this problem setting for real-world environments and suggest a CAPE-Real (CAPER) task setting. This setting allows varied kinematic structures within a semantic category, and multiple instances to co-exist in an observation of real world. To support this task, we build an articulated model repository ReArt-48 and present an efficient dataset generation pipeline, which contains Fast Articulated Object Modeling (FAOM) and Semi-Authentic MixEd Reality Technique (SAMERT). Accompanying the pipeline, we build a large-scale mixed reality dataset ReArtMix and a real world dataset ReArtVal. We also propose an effective framework ReArtNOCS that exploits RGB-D input to estimate part-level pose for multiple instances in a single forward pass. Extensive experiments demonstrate that the proposed ReArtNOCS can achieve good performance on both CAPER and CAPE settings. We believe it could serve as a strong baseline for future research on the CAPER task.