Identifying significant shots in a rally is important for evaluating players' performance in badminton matches. While there are several studies that have quantified player performance in other sports, analyzing badminton data is remained untouched. In this paper, we introduce a badminton language to fully describe the process of the shot and propose a deep learning model composed of a novel short-term extractor and a long-term encoder for capturing a shot-by-shot sequence in a badminton rally by framing the problem as predicting a rally result. Our model incorporates an attention mechanism to enable the transparency of the action sequence to the rally result, which is essential for badminton experts to gain interpretable predictions. Experimental evaluation based on a real-world dataset demonstrates that our proposed model outperforms the strong baselines. The source code is publicly available at https://github.com/yao0510/Shot-Influence.

Measuring event salience is essential in the understanding of stories. This paper takes a recent unsupervised method for salience detection derived from Barthes Cardinal Functions and theories of surprise and applies it to longer narrative forms. We improve the standard transformer language model by incorporating an external knowledgebase (derived from Retrieval Augmented Generation) and adding a memory mechanism to enhance performance on longer works. We use a novel approach to derive salience annotation using chapter-aligned summaries from the Shmoop corpus for classic literary works. Our evaluation against this data demonstrates that our salience detection model improves performance over and above a non-knowledgebase and memory augmented language model, both of which are crucial to this improvement.

A backdoor data poisoning attack is an adversarial attack wherein the attacker injects several watermarked, mislabeled training examples into a training set. The watermark does not impact the test-time performance of the model on typical data; however, the model reliably errs on watermarked examples. To gain a better foundational understanding of backdoor data poisoning attacks, we present a formal theoretical framework within which one can discuss backdoor data poisoning attacks for classification problems. We then use this to analyze important statistical and computational issues surrounding these attacks. On the statistical front, we identify a parameter we call the memorization capacity that captures the intrinsic vulnerability of a learning problem to a backdoor attack. This allows us to argue about the robustness of several natural learning problems to backdoor attacks. Our results favoring the attacker involve presenting explicit constructions of backdoor attacks, and our robustness results show that some natural problem settings cannot yield successful backdoor attacks. From a computational standpoint, we show that under certain assumptions, adversarial training can detect the presence of backdoors in a training set. We then show that under similar assumptions, two closely related problems we call backdoor filtering and robust generalization are nearly equivalent. This implies that it is both asymptotically necessary and sufficient to design algorithms that can identify watermarked examples in the training set in order to obtain a learning algorithm that both generalizes well to unseen data and is robust to backdoors.

Last year, OpenAI introduced the then-largest neural network GPT-3 in a paper titled "Language Models are Few Shot Learners". A state-of-the-art language model, GPT-3, comprises 175 billion parameters against 1.5 billion parameters of its predecessor GPT-2. GPT-3 defeated the Turing NLG model with 17 billion that previously held the record for "largest-ever". The language model has been marvelled at, criticised even, subjected to intense scrutiny; it has found interesting new applications too. All three models have been released within a gap of a year; GPT-1 was released in 2018, GPT-2 in 2019, and GPT-3 in 2020.

Very cool library with lots of great ideas on moving toward'green', efficient multimodal machine learning and AI. Kevin Carlberg, AI Research Science Manager at Facebook Reality Labs (quoted from tweet). PyKale is a PyTorch library for multimodal learning and transfer learning with deep learning and dimensionality reduction on graphs, images, texts, and videos. By adopting a unified pipeline-based API design, PyKale enforces standardization and minimalism, via green machine learning concepts of reducing repetitions and redundancy, reusing existing resources, and recycling learning models across areas. PyKale aims to facilitate interdisciplinary, knowledge-aware machine learning research for graphs, images, texts, and videos in applications including bioinformatics, graph analysis, image/video recognition, and medical imaging.

The idea that software can be developed by artificial intelligence without requiring a human developer opens a world of possibilities -- and questions. Software development AI applications are targeted mainly at developers, promising to act as'co-pilots', and making them more productive. Could this be taken even further to the point where developers are not required at all? What benefit could it have for business users? Having recently been granted preview access to the OpenAI Codex application, Ravi Sawhney took it on a tour through the lens of a business user.

The first barrier is data availability. ML and deep learning models require large datasets to accurately classify or predict different tasks.27 Sectors where ML has seen immense progression are those with large datasets available to enable more complex, precise algorithms.28 In healthcare, however, the availability of data is a complex issue. On the organizational level, health data is not only expensive,27 but there is ingrained reluctance towards data sharing between hospitals as they are considered the property of each hospital to manage their individual patients.29

The Transform Technology Summits start October 13th with Low-Code/No Code: Enabling Enterprise Agility. Deep reinforcement learning is one of the most interesting branches of artificial intelligence. It is behind some of the most remarkable achievements of the AI community, including beating human champions at board and video games, self-driving cars, robotics, and AI hardware design. Deep reinforcement learning leverages the learning capacity of deep neural networks to tackle problems that were too complex for classic RL techniques. Deep reinforcement learning is much more complicated than the other branches of machine learning.

This paper introduces supervised contrastive active learning (SCAL) by leveraging the contrastive loss for active learning in a supervised setting. We propose efficient query strategies in active learning to select unbiased and informative data samples of diverse feature representations. We demonstrate our proposed method reduces sampling bias, achieves state-of-the-art accuracy and model calibration in an active learning setup with the query computation 11x faster than CoreSet and 26x faster than Bayesian active learning by disagreement. Our method yields well-calibrated models even with imbalanced datasets. We also evaluate robustness to dataset shift and out-of-distribution in active learning setup and demonstrate our proposed SCAL method outperforms high performing compute-intensive methods by a bigger margin (average 8.9% higher AUROC for out-of-distribution detection and average 7.2% lower ECE under dataset shift).

COVID-19 frequently provokes pneumonia, which can be diagnosed using imaging exams. Chest X-ray (CXR) is often useful because it is cheap, fast, widespread, and uses less radiation. Here, we demonstrate the impact of lung segmentation in COVID-19 identification using CXR images and evaluate which contents of the image influenced the most. Semantic segmentation was performed using a U-Net CNN architecture, and the classification using three CNN architectures (VGG, ResNet, and Inception). Explainable Artificial Intelligence techniques were employed to estimate the impact of segmentation. A three-classes database was composed: lung opacity (pneumonia), COVID-19, and normal. We assessed the impact of creating a CXR image database from different sources, and the COVID-19 generalization from one source to another. The segmentation achieved a Jaccard distance of 0.034 and a Dice coefficient of 0.982. The classification using segmented images achieved an F1-Score of 0.88 for the multi-class setup, and 0.83 for COVID-19 identification. In the cross-dataset scenario, we obtained an F1-Score of 0.74 and an area under the ROC curve of 0.9 for COVID-19 identification using segmented images. Experiments support the conclusion that even after segmentation, there is a strong bias introduced by underlying factors from different sources.