A Runtime Environment for Contract Automata

By: Davide Basile, Maurice H. ter Beek

Contract automata have been introduced for specifying applications through behavioural contracts and for synthesising their orchestrations as finite state automata. This paper addresses the realisation of applications from contract automata specifications. We present {\tt CARE}, a new runtime environment to coordinate services implementing contracts that guarantees the adherence of the implementation to its contract. We discuss how {\tt... more
Contract automata have been introduced for specifying applications through behavioural contracts and for synthesising their orchestrations as finite state automata. This paper addresses the realisation of applications from contract automata specifications. We present {\tt CARE}, a new runtime environment to coordinate services implementing contracts that guarantees the adherence of the implementation to its contract. We discuss how {\tt CARE} can be adopted to realise contract-based applications, its formal guarantees, and we identify the responsibilities of the involved business actors. Experiments show the benefits of adopting {\tt CARE} with respect to manual implementations. less

DeepCPG Policies for Robot Locomotion

By: Aditya M Deshpande, Eric Hurd, Ali A. Minai, Manish Kumar

Central Pattern Generators (CPGs) form the neural basis of the observed rhythmic behaviors for locomotion in legged animals. The CPG dynamics organized into networks allow the emergence of complex locomotor behaviors. In this work, we take this inspiration for developing walking behaviors in multi-legged robots. We present novel DeepCPG policies that embed CPGs as a layer in a larger neural network and facilitate end-to-end learning of locomo... more
Central Pattern Generators (CPGs) form the neural basis of the observed rhythmic behaviors for locomotion in legged animals. The CPG dynamics organized into networks allow the emergence of complex locomotor behaviors. In this work, we take this inspiration for developing walking behaviors in multi-legged robots. We present novel DeepCPG policies that embed CPGs as a layer in a larger neural network and facilitate end-to-end learning of locomotion behaviors in deep reinforcement learning (DRL) setup. We demonstrate the effectiveness of this approach on physics engine-based insectoid robots. We show that, compared to traditional approaches, DeepCPG policies allow sample-efficient end-to-end learning of effective locomotion strategies even in the case of high-dimensional sensor spaces (vision). We scale the DeepCPG policies using a modular robot configuration and multi-agent DRL. Our results suggest that gradual complexification with embedded priors of these policies in a modular fashion could achieve non-trivial sensor and motor integration on a robot platform. These results also indicate the efficacy of bootstrapping more complex intelligent systems from simpler ones based on biological principles. Finally, we present the experimental results for a proof-of-concept insectoid robot system for which DeepCPG learned policies initially using the simulation engine and these were afterwards transferred to real-world robots without any additional fine-tuning. less

TAU: A Framework for Video-Based Traffic Analytics Leveraging Artificial Intelligence and Unmanned Aerial Systems

By: Bilel Benjdira, Anis Koubaa, Ahmad Taher Azar, Zahid Khan, Adel Ammar, Wadii Boulila

Smart traffic engineering and intelligent transportation services are in increasing demand from governmental authorities to optimize traffic performance and thus reduce energy costs, increase the drivers’ safety and comfort, ensure traffic laws enforcement, and detect traffic violations. In this paper, we address this challenge, and we leverage the use of Artificial Intelligence (AI) and Unmanned Aerial Vehicles (UAVs) to develop an AI-integr... more
Smart traffic engineering and intelligent transportation services are in increasing demand from governmental authorities to optimize traffic performance and thus reduce energy costs, increase the drivers’ safety and comfort, ensure traffic laws enforcement, and detect traffic violations. In this paper, we address this challenge, and we leverage the use of Artificial Intelligence (AI) and Unmanned Aerial Vehicles (UAVs) to develop an AI-integrated video analytics framework, called TAU (Traffic Analysis from UAVs), for automated traffic analytics and understanding. Unlike previous works on traffic video analytics, we propose an automated object detection and tracking pipeline from video processing to advanced traffic understanding using high-resolution UAV images. TAU combines six main contributions. First, it proposes a pre-processing algorithm to adapt the high- resolution UAV image as input to the object detector without lowering the resolution. This ensures an excellent detection accuracy from high-quality features, particularly the small size of detected objects from UAV images. Second, it introduces an algorithm for recalibrating the vehicle coordinates to ensure that vehicles are uniquely identified and tracked across the multiple crops of the same frame. Third, it presents a speed calculation algorithm based on accumulating information from successive frames. Fourth, TAU counts the number of vehicles per traffic zone based on the Ray Tracing algorithm. Fifth, TAU has a fully independent algorithm for crossroad arbitration based on the data gathered from the different zones surrounding it. Sixth, TAU introduces a set of algorithms for extracting twenty-four types of insights from the raw data collected. These insights facilitate the traffic understanding using curves, histograms, heatmaps, and animations. The present work presents a valuable added value for academic researchers and transportation engineers to automate the traffic video analytics process and extract useful insights to optimize traffic performance. TAU is a ready- to-use framework for any Transportation Engineer to better understand and manage daily road traffic (video demonstrations are provided here: https://youtu.be/wXJV0H7LviU and here: https://youtu.be/kGv0gmtVEbI). The source code is available at: https://github.com/bilel-bj/TAU. less

Deep Reinforcement Learning at the Edge of the Statistical Precipice

By: Rishabh Agarwal, Max Schwarzer, Pablo Samuel Castro, Aaron Courville, Marc G. Bellemare

Deep Reinforcement Learning at the Edge of the Statistical Precipice

By: Rishabh Agarwal, Max Schwarzer, Pablo Samuel Castro, Aaron Courville, Marc G. Bellemare

EGRU: Event-based GRU for activity-sparse inference and learning

By: Anand Subramoney, Khaleelulla Khan Nazeer, Mark Schöne, Christian Mayr, David Kappel

The scalability of recurrent neural networks (RNNs) is hindered by the sequential dependence of each time step's computation on the previous time step's output. Therefore, one way to speed up and scale RNNs is to reduce the computation required at each time step independent of model size and task. In this paper, we propose a model that reformulates Gated Recurrent Units (GRU) as an event-based activity-sparse model that we call the Event-base... more
The scalability of recurrent neural networks (RNNs) is hindered by the sequential dependence of each time step's computation on the previous time step's output. Therefore, one way to speed up and scale RNNs is to reduce the computation required at each time step independent of model size and task. In this paper, we propose a model that reformulates Gated Recurrent Units (GRU) as an event-based activity-sparse model that we call the Event-based GRU (EGRU), where units compute updates only on receipt of input events (event-based) from other units. When combined with having only a small fraction of the units active at a time (activity-sparse), this model has the potential to be vastly more compute efficient than current RNNs. Notably, activity-sparsity in our model also translates into sparse parameter updates during gradient descent, extending this compute efficiency to the training phase. We show that the EGRU demonstrates competitive performance compared to state-of-the-art recurrent network models in real-world tasks, including language modeling while maintaining high activity sparsity naturally during inference and training. This sets the stage for the next generation of recurrent networks that are scalable and more suitable for novel neuromorphic hardware. less

NephroNet: A Novel Program for Identifying Renal Cell Carcinoma and Generating Synthetic Training Images with Convolutional Neural Networks and Diffusion Models

By: Yashvir Sabharwal

Renal cell carcinoma (RCC) is a type of cancer that originates in the kidneys and is the most common type of kidney cancer in adults. It can be classified into several subtypes, including clear cell RCC, papillary RCC, and chromophobe RCC. In this study, an artificial intelligence model was developed and trained for classifying different subtypes of RCC using ResNet-18, a convolutional neural network that has been widely used for image classi... more
Renal cell carcinoma (RCC) is a type of cancer that originates in the kidneys and is the most common type of kidney cancer in adults. It can be classified into several subtypes, including clear cell RCC, papillary RCC, and chromophobe RCC. In this study, an artificial intelligence model was developed and trained for classifying different subtypes of RCC using ResNet-18, a convolutional neural network that has been widely used for image classification tasks. The model was trained on a dataset of RCC histopathology images, which consisted of digital images of RCC surgical resection slides that were annotated with the corresponding subtype labels. The performance of the trained model was evaluated using several metrics, including accuracy, precision, and recall. Additionally, in this research, a novel synthetic image generation tool, NephroNet, is developed on diffusion models that are used to generate original images of RCC surgical resection slides. Diffusion models are a class of generative models capable of synthesizing high-quality images from noise. Several diffusers such as Stable Diffusion, Dreambooth Text-to-Image, and Textual Inversion were trained on a dataset of RCC images and were used to generate a series of original images that resembled RCC surgical resection slides, all within the span of fewer than four seconds. The generated images were visually realistic and could be used for creating new training datasets, testing the performance of image analysis algorithms, and training medical professionals. NephroNet is provided as an open-source software package and contains files for data preprocessing, training, and visualization. Overall, this study demonstrates the potential of artificial intelligence and diffusion models for classifying and generating RCC images, respectively. These methods could be useful for improving the diagnosis and treatment of RCC and more. less

Reincarnating Reinforcement Learning: Reusing Prior Computation to Accelerate Progress

By: Rishabh Agarwal, Max Schwarzer, Pablo Samuel Castro, Aaron Courville, Marc G. Bellemare

SoK: Yield Aggregators in DeFi

By: Simon Cousaert, Jiahua Xu, Toshiko Matsui

Yield farming has been an immensely popular activity for cryptocurrency holders since the explosion of Decentralized Finance (DeFi) in the summer of 2020. In this Systematization of Knowledge (SoK), we study a general framework for yield farming strategies with empirical analysis. First, we summarize the fundamentals of yield farming by focusing on the protocols and tokens used by aggregators. We then examine the sources of yield and translat... more
Yield farming has been an immensely popular activity for cryptocurrency holders since the explosion of Decentralized Finance (DeFi) in the summer of 2020. In this Systematization of Knowledge (SoK), we study a general framework for yield farming strategies with empirical analysis. First, we summarize the fundamentals of yield farming by focusing on the protocols and tokens used by aggregators. We then examine the sources of yield and translate those into three example yield farming strategies, followed by the simulations of yield farming performance, based on these strategies. We further compare four major yield aggregrators -- Idle, Pickle, Harvest and Yearn -- in the ecosystem, along with brief introductions of others. We systematize their strategies and revenue models, and conduct an empirical analysis with on-chain data from example vaults, to find a plausible connection between data anomalies and historical events. Finally, we discuss the benefits and risks of yield aggregators. less

Liquidations: DeFi on a Knife-edge

By: Daniel Perez, Sam M. Werner, Jiahua Xu, Benjamin Livshits

The trustless nature of permissionless blockchains renders overcollateralization a key safety component relied upon by decentralized finance (DeFi) protocols. Nonetheless, factors such as price volatility may undermine this mechanism. In order to protect protocols from suffering losses, undercollateralized positions can be liquidated. In this paper, we present the first in-depth empirical analysis of liquidations on protocols for loanable fun... more
The trustless nature of permissionless blockchains renders overcollateralization a key safety component relied upon by decentralized finance (DeFi) protocols. Nonetheless, factors such as price volatility may undermine this mechanism. In order to protect protocols from suffering losses, undercollateralized positions can be liquidated. In this paper, we present the first in-depth empirical analysis of liquidations on protocols for loanable funds (PLFs). We examine Compound, one of the most widely used PLFs, for a period starting from its conception to September 2020. We analyze participants' behavior and risk-appetite in particular, to elucidate recent developments in the dynamics of the protocol. Furthermore, we assess how this has changed with a modification in Compound's incentive structure and show that variations of only 3% in an asset's dollar price can result in over 10m USD becoming liquidable. To further understand the implications of this, we investigate the efficiency of liquidators. We find that liquidators' efficiency has improved significantly over time, with currently over 70% of liquidable positions being immediately liquidated. Lastly, we provide a discussion on how a false sense of security fostered by a misconception of the stability of non-custodial stablecoins, increases the overall liquidation risk faced by Compound participants. less

A Short Survey on Business Models of Decentralized Finance (DeFi) Protocols

By: Teng Andrea Xu, Jiahua Xu

Decentralized Finance (DeFi) services are moving traditional financial operations to the Internet of Value (IOV) by exploiting smart contracts, distributed ledgers, and clever heterogeneous transactions among different protocols. The exponential increase of the Total Value Locked (TVL) in DeFi foreshadows a bright future for automated money transfers in a plethora of services. In this short survey paper, we describe the business model for dif... more
Decentralized Finance (DeFi) services are moving traditional financial operations to the Internet of Value (IOV) by exploiting smart contracts, distributed ledgers, and clever heterogeneous transactions among different protocols. The exponential increase of the Total Value Locked (TVL) in DeFi foreshadows a bright future for automated money transfers in a plethora of services. In this short survey paper, we describe the business model for different DeFi domains - namely, Protocols for Loanable Funds (PLFs), Decentralized Exchanges (DEXs), and Yield Aggregators. We claim that the current state of the literature is still unclear how to value thousands of different competitors (tokens) in DeFi. With this work, we abstract the general business model for different DeFi domains and compare them. Finally, we provide open research challenges that will involve heterogeneous domains such as economics, finance, and computer science. less

The Anatomy of a Cryptocurrency Pump-and-Dump Scheme

By: Jiahua Xu, Benjamin Livshits

While pump-and-dump schemes have attracted the attention of cryptocurrency observers and regulators alike, this paper represents the first detailed empirical query of pump-and-dump activities in cryptocurrency markets. We present a case study of a recent pump-and-dump event, investigate 412 pump-and-dump activities organized in Telegram channels from June 17, 2018 to February 26, 2019, and discover patterns in crypto-markets associated with p... more
While pump-and-dump schemes have attracted the attention of cryptocurrency observers and regulators alike, this paper represents the first detailed empirical query of pump-and-dump activities in cryptocurrency markets. We present a case study of a recent pump-and-dump event, investigate 412 pump-and-dump activities organized in Telegram channels from June 17, 2018 to February 26, 2019, and discover patterns in crypto-markets associated with pump-and-dump schemes. We then build a model that predicts the pump likelihood of all coins listed in a crypto-exchange prior to a pump. The model exhibits high precision as well as robustness, and can be used to create a simple, yet very effective trading strategy, which we empirically demonstrate can generate a return as high as 60% on small retail investments within a span of two and half months. The study provides a proof of concept for strategic crypto-trading and sheds light on the application of machine learning for crime detection. less

SoK: Decentralized Exchanges (DEX) with Automated Market Maker (AMM) Protocols

By: Jiahua Xu, Krzysztof Paruch, Simon Cousaert, Yebo Feng

As an integral part of the decentralized finance (DeFi) ecosystem, decentralized exchanges (DEXs) with automated market maker (AMM) protocols have gained massive traction with the recently revived interest in blockchain and distributed ledger technology (DLT) in general. Instead of matching the buy and sell sides, automated market makers (AMMs) employ a peer-to-pool method and determine asset price algorithmically through a so-called conserva... more
As an integral part of the decentralized finance (DeFi) ecosystem, decentralized exchanges (DEXs) with automated market maker (AMM) protocols have gained massive traction with the recently revived interest in blockchain and distributed ledger technology (DLT) in general. Instead of matching the buy and sell sides, automated market makers (AMMs) employ a peer-to-pool method and determine asset price algorithmically through a so-called conservation function. To facilitate the improvement and development of automated market maker (AMM)-based decentralized exchanges (DEXs), we create the first systematization of knowledge in this area. We first establish a general automated market maker (AMM) framework describing the economics and formalizing the system's state-space representation. We then employ our framework to systematically compare the top automated market maker (AMM) protocols' mechanics, illustrating their conservation functions, as well as slippage and divergence loss functions. We further discuss security and privacy concerns, how they are enabled by automated market maker (AMM)-based decentralized exchanges (DEXs)' inherent properties, and explore mitigating solutions. Finally, we conduct a comprehensive literature review on related work covering both decentralized finance (DeFi) and conventional market microstructure. less

The Augmentation-Speed Tradeoff for Consistent Network Updates

By: Monika Henzinger, Ami Paz, Arash Pourdamghani, Stefan Schmid

Emerging software-defined networking technologies enable more adaptive communication infrastructures, allowing for quick reactions to changes in networking requirements by exploiting the workload's temporal structure. However, operating networks adaptively is algorithmically challenging, as meeting networks' stringent dependability requirements relies on maintaining basic consistency and performance properties, such as loop freedom and conges... more
Emerging software-defined networking technologies enable more adaptive communication infrastructures, allowing for quick reactions to changes in networking requirements by exploiting the workload's temporal structure. However, operating networks adaptively is algorithmically challenging, as meeting networks' stringent dependability requirements relies on maintaining basic consistency and performance properties, such as loop freedom and congestion minimization, even during the update process. This paper leverages an augmentation-speed tradeoff to significantly speed up consistent network updates. We show that allowing for a small and short (hence practically tolerable, e.g., using buffering) oversubscription of links allows us to solve many network update instances much faster, as well as to reduce computational complexities (i.e., the running times of the algorithms). We first explore this tradeoff formally, revealing the computational complexity of scheduling updates. We then present and analyze algorithms that maintain logical and performance properties during the update. Using an extensive simulation study, we find that the tradeoff is even more favorable in practice than our analytical bounds suggest. In particular, we find that by allowing just 10% augmentation, update times reduce by more than 32% on average, across a spectrum of real-world networks. less

Reposing Humans by Warping 3D Features

By: Markus Knoche, István Sárándi, Bastian Leibe

We address the problem of reposing an image of a human into any desired novel pose. This conditional image-generation task requires reasoning about the 3D structure of the human, including self-occluded body parts. Most prior works are either based on 2D representations or require fitting and manipulating an explicit 3D body mesh. Based on the recent success in deep learning-based volumetric representations, we propose to implicitly learn a d... more
We address the problem of reposing an image of a human into any desired novel pose. This conditional image-generation task requires reasoning about the 3D structure of the human, including self-occluded body parts. Most prior works are either based on 2D representations or require fitting and manipulating an explicit 3D body mesh. Based on the recent success in deep learning-based volumetric representations, we propose to implicitly learn a dense feature volume from human images, which lends itself to simple and intuitive manipulation through explicit geometric warping. Once the latent feature volume is warped according to the desired pose change, the volume is mapped back to RGB space by a convolutional decoder. Our state-of-the-art results on the DeepFashion and the iPER benchmarks indicate that dense volumetric human representations are worth investigating in more detail. less

MeTRAbs: Metric-Scale Truncation-Robust Heatmaps for Absolute 3D Human Pose Estimation

By: István Sárándi, Timm Linder, Kai O. Arras, Bastian Leibe

Heatmap representations have formed the basis of human pose estimation systems for many years, and their extension to 3D has been a fruitful line of recent research. This includes 2.5D volumetric heatmaps, whose X and Y axes correspond to image space and Z to metric depth around the subject. To obtain metric-scale predictions, 2.5D methods need a separate post-processing step to resolve scale ambiguity. Further, they cannot localize body join... more
Heatmap representations have formed the basis of human pose estimation systems for many years, and their extension to 3D has been a fruitful line of recent research. This includes 2.5D volumetric heatmaps, whose X and Y axes correspond to image space and Z to metric depth around the subject. To obtain metric-scale predictions, 2.5D methods need a separate post-processing step to resolve scale ambiguity. Further, they cannot localize body joints outside the image boundaries, leading to incomplete estimates for truncated images. To address these limitations, we propose metric-scale truncation-robust (MeTRo) volumetric heatmaps, whose dimensions are all defined in metric 3D space, instead of being aligned with image space. This reinterpretation of heatmap dimensions allows us to directly estimate complete, metric-scale poses without test-time knowledge of distance or relying on anthropometric heuristics, such as bone lengths. To further demonstrate the utility our representation, we present a differentiable combination of our 3D metric-scale heatmaps with 2D image-space ones to estimate absolute 3D pose (our MeTRAbs architecture). We find that supervision via absolute pose loss is crucial for accurate non-root-relative localization. Using a ResNet-50 backbone without further learned layers, we obtain state-of-the-art results on Human3.6M, MPI-INF-3DHP and MuPoTS-3D. Our code will be made publicly available to facilitate further research. less

Learning 3D Human Pose Estimation from Dozens of Datasets using a Geometry-Aware Autoencoder to Bridge Between Skeleton Formats

By: István Sárándi, Alexander Hermans, Bastian Leibe

Deep learning-based 3D human pose estimation performs best when trained on large amounts of labeled data, making combined learning from many datasets an important research direction. One obstacle to this endeavor are the different skeleton formats provided by different datasets, i.e., they do not label the same set of anatomical landmarks. There is little prior research on how to best supervise one model with such discrepant labels. We show t... more
Deep learning-based 3D human pose estimation performs best when trained on large amounts of labeled data, making combined learning from many datasets an important research direction. One obstacle to this endeavor are the different skeleton formats provided by different datasets, i.e., they do not label the same set of anatomical landmarks. There is little prior research on how to best supervise one model with such discrepant labels. We show that simply using separate output heads for different skeletons results in inconsistent depth estimates and insufficient information sharing across skeletons. As a remedy, we propose a novel affine-combining autoencoder (ACAE) method to perform dimensionality reduction on the number of landmarks. The discovered latent 3D points capture the redundancy among skeletons, enabling enhanced information sharing when used for consistency regularization. Our approach scales to an extreme multi-dataset regime, where we use 28 3D human pose datasets to supervise one model, which outperforms prior work on a range of benchmarks, including the challenging 3D Poses in the Wild (3DPW) dataset. Our code and models are available for research purposes. less

Dynamic Membership for Regular Languages

By: Antoine Amarilli, Louis Jachiet, Charles Paperman

We study the dynamic membership problem for regular languages: fix a language L, read a word w, build in time O(|w|) a data structure indicating if w is in L, and maintain this structure efficiently under letter substitutions on w. We consider this problem on the unit cost RAM model with logarithmic word length, where the problem always has a solution in O(log |w| / log log |w|) per operation. We show that the problem is in O(log log |w|) for... more
We study the dynamic membership problem for regular languages: fix a language L, read a word w, build in time O(|w|) a data structure indicating if w is in L, and maintain this structure efficiently under letter substitutions on w. We consider this problem on the unit cost RAM model with logarithmic word length, where the problem always has a solution in O(log |w| / log log |w|) per operation. We show that the problem is in O(log log |w|) for languages in an algebraically-defined, decidable class QSG, and that it is in O(1) for another such class QLZG. We show that languages not in QSG admit a reduction from the prefix problem for a cyclic group, so that they require {\Omega}(log |w| / log log |w|) operations in the worst case; and that QSG languages not in QLZG admit a reduction from the prefix problem for the multiplicative monoid U 1 = {0, 1}, which we conjecture cannot be maintained in O(1). This yields a conditional trichotomy. We also investigate intermediate cases between O(1) and O(log log |w|). Our results are shown via the dynamic word problem for monoids and semigroups, for which we also give a classification. We thus solve open problems of the paper of Skovbjerg Frandsen, Miltersen, and Skyum [30] on the dynamic word problem, and additionally cover regular languages. less

Surrogate "Level-Based" Lagrangian Relaxation for Mixed-Integer Linear Programming

By: Mikhail A. Bragin, Emily L. Tucker

Mixed-Integer Linear Programming (MILP) plays an important role across a range of scientific disciplines and within areas of strategic importance to society. The MILP problems, however, suffer from combinatorial complexity. Because of integer decision variables, as the problem size increases, the number of possible solutions increases super-linearly thereby leading to a drastic increase in the computational effort. To efficiently solve MILP p... more
Mixed-Integer Linear Programming (MILP) plays an important role across a range of scientific disciplines and within areas of strategic importance to society. The MILP problems, however, suffer from combinatorial complexity. Because of integer decision variables, as the problem size increases, the number of possible solutions increases super-linearly thereby leading to a drastic increase in the computational effort. To efficiently solve MILP problems, a "price-based" decomposition and coordination approach is developed to exploit 1. the super-linear reduction of complexity upon the decomposition and 2. the geometric convergence potential inherent to Polyak's stepsizing formula for the fastest coordination possible to obtain near-optimal solutions in a computationally efficient manner. Unlike all previous methods to set stepsizes heuristically by adjusting hyperparameters, the key novel way to obtain stepsizes is purely decision-based: a novel "auxiliary" constraint satisfaction problem is solved, from which the appropriate stepsizes are inferred. Testing results for large-scale Generalized Assignment Problems (GAP) demonstrate that for the majority of instances, certifiably optimal solutions are obtained. For stochastic job-shop scheduling as well as for pharmaceutical scheduling, computational results demonstrate the two orders of magnitude speedup as compared to Branch-and-Cut (B&C). The new method has a major impact on the efficient resolution of complex Mixed-Integer Programming (MIP) problems arising within a variety of scientific fields. less

Multi-Label Chest X-Ray Classification via Deep Learning

By: Aravind Sasidharan Pillai

In this era of pandemic, the future of healthcare industry has never been more exciting. Artificial intelligence and machine learning (AI & ML) present opportunities to develop solutions that cater for very specific needs within the industry. Deep learning in healthcare had become incredibly powerful for supporting clinics and in transforming patient care in general. Deep learning is increasingly being applied for the detection of clinically ... more
In this era of pandemic, the future of healthcare industry has never been more exciting. Artificial intelligence and machine learning (AI & ML) present opportunities to develop solutions that cater for very specific needs within the industry. Deep learning in healthcare had become incredibly powerful for supporting clinics and in transforming patient care in general. Deep learning is increasingly being applied for the detection of clinically important features in the images beyond what can be perceived by the naked human eye. Chest X-ray images are one of the most common clinical method for diagnosing a number of diseases such as pneumonia, lung cancer and many other abnormalities like lesions and fractures. Proper diagnosis of a disease from X-ray images is often challenging task for even expert radiologists and there is a growing need for computerized support systems due to the large amount of information encoded in X-Ray images. The goal of this paper is to develop a lightweight solution to detect 14 different chest conditions from an X ray image. Given an X-ray image as input, our classifier outputs a label vector indicating which of 14 disease classes does the image fall into. Along with the image features, we are also going to use non-image features available in the data such as X-ray view type, age, gender etc. The original study conducted Stanford ML Group is our base line. Original study focuses on predicting 5 diseases. Our aim is to improve upon previous work, expand prediction to 14 diseases and provide insight for future chest radiography research. less