Papers

We consider the car key localization task using ultra-wideband (UWB) signal measurements. Given labeled data for a certain car, we train a deep classifier to make the prediction about the new points. However, due to the differences in car models and possible environmental effects that might alter the signal propagation, data collection requires considerable effort for each car. In particular, we consider a situation where the data for the new car is collected only in one environment, so we have to utilize the measurements in other environments from a different car. We propose a framework based on generative adversarial networks (GANs) to generate missing parts of the data and train the classifier on it, mitigating the necessity to collect the real data. We show that the model trained on the synthetic data performs better than the baseline trained on the collected measurements only. Furthermore, our model closes the gap to the level of performance achieved when we would have the information about the new car in multiple environments by 35%.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

The doctoral work summarized here is an application of Artificial Intelligence (AI) for social good. The successful implementation would contribute towards low-cost, faster monitoring of road conditions across different nations, resulting in safer roads for everyone. Additionally, the study provides recommendations for re-using the road image data and the Deep Learning models released by any country for detecting road damage in other countries.

poster session 3 @ red 6, poster session 1 @ red 6, poster session 3, poster session 1

Implementations of artificial intelligence (AI) based on deep learning (DL) have proven to be highly successful in many domains, from biomedical imaging to natural language processing, but are still rarely applied in the space industry, particularly for onboard learning of planetary surfaces. In this project, we discuss the utility and limitations of DL, enhanced with topological footprints of the sensed objects, for multi-class classification of planetary surface patterns, in conjunction with tactile and embedded sensing in rover exploratory missions. We consider a Topological Convolutional Network (TCN) model with a persistence-based attention mechanism for supervised classification of various landforms. We study TCN's performance on the Barefoot surface pattern dataset, a novel surface pressure dataset from a prototype tactile rover wheel, known as the Barefoot Rover tactile wheel. Multi-class pattern recognition in the Barefoot data has neither been ever tackled before with DL nor assessed with topological methods. We provide insights into advantages and restrictions of topological DL as the early-stage concept for onboard learning and planetary exploration.

poster session 1 @ red 6, poster session 9 @ red 6, poster session 1, poster session 9

In the past decade, automatic product description generation for e-commerce have witnessed significant advancement. As the services provided by e-commerce platforms become diverse, it is necessary to dynamically adapt the patterns of descriptions generated. The selling point of products is an important type of product description for which the length should be as short as possible while still conveying key information. In addition, this kind of product description should be eye-catching to the readers. Currently, product selling points are normally written by human experts. Thus, the creation and maintenance of these contents incur high costs. These costs can be significantly reduced if product selling points can be automatically generated by machines. In this paper, we report our experience developing and deploying the Intelligent Online Selling Point Extraction (IOSPE) system to serve the recommendation system in the JD.com e-commerce platform. Since July 2020, IOSPE has become a core service for 62 key categories of products (covering more than 4 million products). So far, it has generated more than 1.1 billion selling points, thereby significantly scaling up the selling point creation operation and saving human labour. These IOSPE generated selling points have increased the click-through rate (CTR) by 1.89% and the average duration the customers spent on the products by more than 2.03% compared to the previous practice, which are significant improvements for such a large-scale e-commerce platform.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

Breast reconstruction surgery requires extensive planning, usually with a CT scan that helps surgeons identify which vessels are suitable for harvest. Currently, there is no quantitative method for preoperative planning. In this work, we successfully develop a Deep Learning algorithm to segment the vessels within the region of interest for breast reconstruction. Ultimately, this information will be used to determine the optimal reconstructive method (choice of vessels, extent of the free flap/harvested tissue) to reduce intra- and postoperative complication rates.

poster session 3 @ red 6, poster session 1 @ red 6, poster session 3, poster session 1

Intensive care in hospitals is distributed to different units that care for patient populations reflecting specific comorbidities, treatments, and outcomes. Unit expertise can be shared to potentially improve the quality of methods and outcomes for patients across units. We propose an algorithmic rule pruning approach for use in building short lists of human-interpretable rules that reliably identify patient beneficiaries of expertise transfers in the form of machine learning risk models. Our experimental results, obtained with two intensive care monitoring datasets, demonstrate the potential utility of the proposed method in practice.

poster session 1 @ red 6, poster session 2 @ red 6, poster session 1, poster session 2

As climate change is increasing the frequency and intensity of climate and weather hazards, improving detection and monitoring of flood events is a priority. Being weather independent and high resolution, Sentinel 1 (S1) radar satellite imagery data has become the go to data source to detect flood events accurately. However, current methods are either based on fixed thresholds to differentiate water from land or train Artificial Intelligence (AI) models based on only S1 data, despite the availability of many other relevant data sources publicly. These models also lack comprehensive validations on out-of-sample data and deployment at scale. In this study, we investigated whether adding extra input layers could increase the performance of AI models in detecting floods from S1 data. We also provide performance across a range of 11 historical events, with results ranging between 0.93 and 0.97 accuracy, 0.53 and 0.81 IoU, and 0.68 and 0.89 F1 scores. Finally, we show the infrastructure we developed to deploy our AI models at scale to satisfy a range of use cases and user requests.

poster session 1 @ red 6, poster session 9 @ red 6, poster session 1, poster session 9

With increasing world population and expanded use of forests as cohabited regions, interactions and conflicts with wildlife are increasing, leading to large scale loss of lives (animal and human) and livelihoods (economic). While community knowledge is valuable, forest officials and conservation organisations can greatly benefit from predictive analysis of human-wildlife conflict, leading to targeted interventions that can potentially help save lives and livelihoods. However, the problem of prediction is a complex socio-technical problem in the context of limited data in low-resource regions. Identifying the right features to make accurate predictions of conflicts at the required spatial granularity using a sparse conflict training dataset is the key challenge that we address in this paper. Specifically, we do an illustrative case study on human-wildlife conflicts in the Bramhapuri Forest Division in Chandrapur, Maharashtra, India. Most existing work has considered human wildlife conflicts in protected areas and to the best of our knowledge, this is the first effort at prediction of human-wildlife conflicts in unprotected areas and using those predictions for deploying interventions on the ground.

poster session 1 @ red 6, poster session 9 @ red 6, poster session 1, poster session 9

The CO2 capture efficiency in solvent-based carbon capture systems (CCSs) critically depends on the gas-solvent interfacial area (IA), making maximization of IA a foundational challenge in CCS design. While the IA associated with a particular CCS design can be estimated via a computational fluid dynamics (CFD) simulation, using CFD to derive the IAs associated with numerous CCS designs is prohibitively costly. Fortunately, previous works such as Deep Fluids (DF) (Kim et al., 2019) show that large simulation speedups are achievable by replacing CFD simulators with neural network (NN) surrogates that faithfully mimic the CFD simulation process. This raises the possibility of a fast, accurate replacement for a CFD simulator and therefore efficient approximation of the IAs required by CCS design optimization. Thus, here, we build on the DF approach to develop surrogates that can successfully be applied to our complex carbon-capture CFD simulations. Our optimized DF-style surrogates produce large speedups (4000x) while obtaining IA relative errors as low as 4% on unseen CCS configurations that lie within the range of training configurations. This hints at the promise of NN surrogates for our CCS design optimization problem. Nonetheless, DF has inherent limitations with respect to CCS design (e.g., limited transferability of trained models to new CCS packings). We conclude with ideas to address these challenges.

poster session 1 @ red 6, poster session 9 @ red 6, poster session 1, poster session 9

Item representation learning is crucial for search and recommendation tasks in e-commerce. In e-commerce, the instances (e.g., items, users) in different domains are always related. Such instance relationship across domains contains useful local information for transfer learning. However, existing transfer learning based approaches did not leverage this knowledge. In this paper, we report on our experience designing and deploying Prior-Guided Transfer Learning (PGTL) to bridge this gap. It utilizes the instance relationship across domains to extract prior knowledge for the target domain and leverages it to guide the fine-grained transfer learning for e-commerce item representation learning tasks. Rather than directly transferring knowledge from the source domain to the target domain, the prior knowledge can serve as a bridge to link both domains and enhance knowledge transfer, especially when the domain distribution discrepancy is large. Since its deployment on the Taiwanese portal of Taobao in Aug 2020, PGTL has significantly improved the item exposure rate and item click-through rate compared to previous approaches

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

Product copywriting is a critical component of e-commerce recommendation platforms. It aims to attract users' interest and improve user experience by highlighting product characteristics with textual descriptions. In this paper, we report our experience deploying the proposed Automatic Product Copywriting Generation (APCG) system into the JD.com e-commerce product recommendation platform. It consists of two main components: 1) natural language generation, which is built from a transformer-pointer network and a pre-trained sequence-to-sequence model based on millions of training data from our in-house platform; and 2) copywriting quality control, which is based on both automatic evaluation and human screening. For selected domains, the models are trained and updated daily with the updated training data. In addition, the model is also used as a real-time writing assistant tool on our live broadcast platform. The APCG system has been deployed in JD.com since Feb 2021. By Sep 2021, it has generated 2.53 million product descriptions, and improved the overall averaged click-through rate (CTR) and the Conversion Rate (CVR) by 4.22% and 3.61%, compared to baselines, respectively on a year-on-year basis. The accumulated Gross Merchandise Volume (GMV) made by our system is improved by 213.42%, compared to the number in Feb 2021.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

African smallholder farmers have struggled with low agricultural productivity for decades, partly due to their inability to proactively assess irrigation needs in their farms in the face of long-term climate change. In this paper, we tackle this challenge by employing data-driven techniques to develop forecasting tools for three widely used crop-productivity related variables (i.e., actual evapotranspiration, reference evapotranspiration, and net primary production), which can then be used by farmers to take corrective actions on their farms. Prior work in this domain, despite using data-driven methods, suffers from two major limitations: (i) they mainly focus on estimating variable values (as opposed to forecasting the future); and (ii) they mostly use classical Machine Learning (ML) prediction models, despite the abundance of data sufficient to train sophisticated deep learning models. To fill this research gap, we collaborate with PlantVillage, the world’s leading non-profit agricultural knowledge delivery platform for African farmers, to identify ∼2,200 smallholder farm locations, and gather remote-sensed data of these farms over a period of five years. Next, we propose CLIMATES, a meta-algorithm leveraging structural insights about temporal patterns of this time-series data to accurately forecast their future values. We conduct extensive experiments to evaluate its performance in this domain. Our experimental results show that CLIMATES outperforms several state-of-the-art time-series forecasting models. We also provide insights about the poor performance of some competing models. Our work is being evaluated by officials at PlantVillage for potential future deployment as an early warning system in East Africa. We release the code at https://github.com/maryam-tabar/CLIMATES.

poster session 1 @ red 6, poster session 9 @ red 6, poster session 1, poster session 9

Web search engines focus on serving highly relevant results within hundreds of milliseconds. Pre-trained language transformer models such as BERT are therefore hard to use in this scenario due to their high computational demands. We present our real-time approach to the document ranking problem leveraging a BERT-based siamese architecture. The model is already deployed in a commercial search engine and it improves production performance by more than 3%. For further research and evaluation, we release DaReCzech, a unique data set of 1.6 million Czech user query-document pairs with manually assigned relevance levels. We also release Small-E-Czech, an Electra-small language model pre-trained on a large Czech corpus. We believe this data will support endeavours both of search relevance and multilingual-focused research communities.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

More than US$ 27 billion is estimated to have been paid-out in farm support in USA alone since 1991 in response to climate change impacts on agriculture, with costs likely continuing to rise. With the wider adoption of precision agriculture - an agriculture management strategy that involves gathering, processing and analyzing temporal, spatial and individual data - in both developed and developing countries, there is an increasing opportunity to harness accumulating, shareable, big data using artificial intelligence (AI) methods, collected from weather stations, field sensor networks, Internet-of-Things devices, unmanned aerial vehicles, and earth observational satellites. This requires smart algorithms tailored to agricultural data types, integrated into digital solutions that are viable, flexible, and scalable for wide deployment for a wide variety of agricultural users and decision-makers. We discuss a novel AI approach that addresses the real-world problem of developing a viable solution for reliably, timely, and cost-effectively forecasting crop status across large agricultural regions using Earth observational information in near-real-time. Our approach is based on extracting time-conditioned topological features which characterize complex spatio-temporal dependencies between crop production regions and integrating such topological signatures into Long Short Term Memory (LSTM). We discuss utility and limitations of the resulting zigzag persistence-based LSTM (ZZTop-LSTM) as a new tool for developing more informed crop insurance rate-making and accurate tracking of changing risk exposures and vulnerabilities within insurance risk areas.

poster session 1 @ red 6, poster session 9 @ red 6, poster session 1, poster session 9

Product allocation in retail is the process of placing products throughout a store to connect consumers with relevant products. Discovering a good allocation strategy is challenging due to the scarcity of data and the high cost of experimentation in the physical world. Some work explores Reinforcement learning (RL) as a solution, but these approaches are often limited because of the sim2real problem. Learning policies from logged trajectories of a system is a key step forward for RL in physical systems. Recent work has shown that model-based offline RL can improve the effectiveness of offline policy estimation through uncertainty-penalized exploration. However, existing work assumes a continuous state space and access to a covariance matrix of the environment dynamics, which is not possible in the discrete case. To solve this problem, we propose a Bayesian model-based technique that naturally produces probabilistic estimates of the environment dynamics via the posterior predictive distribution, which we use for uncertainty-penalized exploration. We call our approach Posterior Penalized Offline Policy Optimization (PPOPO). We show that our world model better fits historical data due to informative priors, and that PPOPO outperforms other offline techniques in simulation and against real-world data.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

Classical measurements and modelling that underpin present flood warning and alert systems are based on fixed and spatially restricted static sensor networks. Computationally expensive physics-based simulations are often used that can't react in real-time to changes in environmental conditions. We want to explore contemporary artificial intelligence (AI) for predicting flood risk in real time by using a diverse range of data sources. By combining heterogeneous data sources, we aim to nowcast rapidly changing flood conditions and gain a grater understanding of urgent humanitarian needs.

poster session 3 @ red 6, poster session 1 @ red 6, poster session 3, poster session 1

Driver's anxiety about the remaining driving range of electric vehicles (EVs) has been quite improved by mounting a high-capacity battery pack. However, when EVs need to be charged, the drivers still feel uncomfortable if inaccurate range prediction is provided because the inaccuracy makes it difficult to decide when and where to charge EV. In this paper, to mitigate the EV range anxiety, a new machine learning (ML) method to enhance range prediction accuracy is proposed in a practical way. For continuously obtaining the recent traffic conditions ahead, input features indicating the near-future vehicle dynamics are connected to a long short-term memory (LSTM) network, which can consecutively utilize a relation of neighboring data, and then the output features of the LSTM network with another input features consisting of energy-related vehicle system states become another input layer for deep learning network (DNN). The proposed LSTM-DNN mixture model is trained by exploiting the driving data of about 160,000 km and the following test performance shows that the model retains the range prediction accuracy of 2 ~ 3 km in a time window of 40 min. The test results indicate that the LSTM-DNN range prediction model is able to make a far-sighted range prediction while considering varying map and traffic information to a destination.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8

In this paper, we address a crucial problem in fashion e-commerce (with respect to customer experience, as well as revenue): color variants identification, i.e., identifying fashion products that match exactly in their design (or style), but only to differ in their color. We propose a generic framework, that leverages deep visual Representation Learning at its heart, to address this problem for our fashion e-commerce platform. Our framework could be trained with supervisory signals in the form of triplets, that are obtained manually. However, it is infeasible to obtain manual annotations for the entire huge collection of data usually present in fashion e-commerce platforms, such as ours, while capturing all the difficult corner cases. But, to our rescue, interestingly we observed that this crucial problem in fashion e-commerce could also be solved by simple color jitter based image augmentation, that recently became widely popular in the contrastive Self-Supervised Learning (SSL) literature, that seeks to learn visual representations without using manual labels. This naturally led to a question in our mind: Could we leverage SSL in our use-case, and still obtain comparable performance to our supervised framework? The answer is, Yes! because, color variant fashion objects are nothing but manifestations of a style, in different colors, and a model trained to be invariant to the color (with, or without supervision), should be able to recognize this! This is what the paper further demonstrates, both qualitatively, and quantitatively, while evaluating a couple of state-of-the-art SSL techniques, and also proposing a novel method.

poster session 1 @ red 6, poster session 8 @ red 6, poster session 1, poster session 8