While federated learning (FL) has recently emerged as a promising approach to train machine learning models, it is limited to only preliminary explorations in the domain of automatic speech recognition (ASR). Moreover, FL does not inherently guarantee user privacy and requires the use of differential privacy (DP) for robust privacy guarantees. However, we are not aware of prior work on applying DP to FL for ASR. In this paper, we aim to bridge this research gap by formulating an ASR benchmark for FL with DP and establishing the first baselines. First, we extend the existing research on FL for ASR by exploring different aspects of recent $\textit{large end-to-end transformer models}$: architecture design, seed models, data heterogeneity, domain shift, and impact of cohort size. With a $\textit{practical}$ number of central aggregations we are able to train $\textbf{FL models}$ that are \textbf{nearly optimal} even with heterogeneous data, a seed model from another domain, or no pre-trained seed model. Second, we apply DP to FL for ASR, which is non-trivial since DP noise severely affects model training, especially for large transformer models, due to highly imbalanced gradients in the attention block. We counteract the adverse effect of DP noise by reviving per-layer clipping and explaining why its effect is more apparent in our case than in the prior work. Remarkably, we achieve user-level ($7.2$, $10^{-9}$)-$\textbf{DP}$ (resp. ($4.5$, $10^{-9}$)-$\textbf{DP}$) with a 1.3% (resp. 4.6%) absolute drop in the word error rate for extrapolation to high (resp. low) population scale for $\textbf{FL with DP in ASR}$.

Audio-visual speech contains synchronized audio and visual information that provides cross-modal supervision to learn representations for both automatic speech recognition (ASR) and visual speech recognition (VSR). We introduce continuous pseudo-labeling for audio-visual speech recognition (AV-CPL), a semi-supervised method to train an audio-visual speech recognition (AVSR) model on a combination of labeled and unlabeled videos with continuously regenerated pseudo-labels. Our models are trained for speech recognition from audio-visual inputs and can perform speech recognition using both audio and visual modalities, or only one modality. Our method uses the same audio-visual model for both supervised training and pseudo-label generation, mitigating the need for external speech recognition models to generate pseudo-labels. Finally, using visual-only speech data, our method is able to leverage unlabeled visual speech to improve VSR. Machine learning has enabled rapid advancement in fields such as speech processing. However, speech processing requires large amounts of labeled data to work well (Radford et al., 2023; Zheng et al., 2022), which is hard to acquire for the thousands of languages spoken world-wide. Semisupervised learning aims to mitigate this challenge by using unlabeled data to learn better representations and improve performance on labeled data. Real-world unlabeled data is often multi-modal, for example, videos containing synchronized audio and visual information. In this work, we investigate whether we can use such multi-modal data in a semi-supervised pipeline to improve performance on labeled data. Multi-modal data has an additional benefit - modalities can be complementary for each other and provide cross-modal supervision, which influences our algorithm design. In this work, we study audio-visual speech as multi-modal data with synchronized audio and visual input sequences. Using only the audio or the video data, we can perform two kinds of speech recognition: automatic speech recognition (ASR) from the audio channel, or visual speech recognition (VSR) from the video channel (lip-reading). However, these modalities require substantially different amounts of labeled data for training practical models. For example, with 30 hours of labeled data, we can train an ASR model which reaches around 11% word error rate (WER), while training modern end-to-end VSR models on the same amount of data is challenging: the lowest WER we achieve in our experiments is 96%.

Information value (IV) is a quite popular technique for features selection before the modeling phase. There are practical criteria, based on fixed thresholds for IV, but at the same time mysterious and lacking theoretical arguments, to decide if a predictor has sufficient predictive power to be considered in the modeling phase. However, the mathematical development and statistical inference methods for this technique are almost nonexistent in the literature. In this paper we present a theoretical framework for IV, and at the same time, we propose a non-parametric hypothesis test to evaluate the predictive power of features contemplated in a data set. Due to its relationship with divergence measures developed in the Information Theory, we call our proposal the J - Divergence test. We show how to efficiently compute our test statistic and we study its performance on simulated data. In various scenarios, particularly in unbalanced data sets, we show its superiority over conventional criteria based on fixed thresholds. Furthermore, we apply our test on fraud identification data and provide an open-source Python library, called "statistical-iv"(https://pypi.org/project/statistical-iv/), where we implement our main results.

Graph unlearning has emerged as a pivotal method to delete information from a pre-trained graph neural network (GNN). One may delete nodes, a class of nodes, edges, or a class of edges. An unlearning method enables the GNN model to comply with data protection regulations (i.e., the right to be forgotten), adapt to evolving data distributions, and reduce the GPU-hours carbon footprint by avoiding repetitive retraining. Existing partitioning and aggregation-based methods have limitations due to their poor handling of local graph dependencies and additional overhead costs. More recently, GNNDelete offered a model-agnostic approach that alleviates some of these issues. Our work takes a novel approach to address these challenges in graph unlearning through knowledge distillation, as it distills to delete in GNN (D2DGN). It is a model-agnostic distillation framework where the complete graph knowledge is divided and marked for retention and deletion. It performs distillation with response-based soft targets and feature-based node embedding while minimizing KL divergence. The unlearned model effectively removes the influence of deleted graph elements while preserving knowledge about the retained graph elements. D2DGN surpasses the performance of existing methods when evaluated on various real-world graph datasets by up to $43.1\%$ (AUC) in edge and node unlearning tasks. Other notable advantages include better efficiency, better performance in removing target elements, preservation of performance for the retained elements, and zero overhead costs. Notably, our D2DGN surpasses the state-of-the-art GNNDelete in AUC by $2.4\%$, improves membership inference ratio by $+1.3$, requires $10.2\times10^6$ fewer FLOPs per forward pass and up to $\mathbf{3.2}\times$ faster.

In this task, given an image and a textual question, one has to draw the bounding box around the object correctly responding to that question. Every image-question pair contains the response, with only one correct response per image. Our dataset contains 45,199 pairs of images and questions in English, provided with ground truth bounding boxes, split into train and two test subsets. Besides describing the dataset and releasing it under a CC BY license, we conducted a series of experiments on open source zero-shot baseline models and organized a multi-phase competition at WSDM Cup that attracted 48 participants worldwide. However, by the time of paper submission, no machine learning model outperformed the non-expert crowdsourcing baseline according to the intersection over union evaluation score.

Searching for available parking spots in high-density urban centers is a stressful task for drivers that can be mitigated by systems that know in advance the nearest parking space available. To this end, image-based systems offer cost advantages over other sensor-based alternatives (e.g., ultrasonic sensors), requiring less physical infrastructure for installation and maintenance. Despite recent deep learning advances, deploying intelligent parking monitoring is still a challenge since most approaches involve collecting and labeling large amounts of data, which is laborious and time-consuming. Our study aims to uncover the challenges in creating a global framework, trained using publicly available labeled parking lot images, that performs accurately across diverse scenarios, enabling the parking space monitoring as a ready-to-use system to deploy in a new environment. Through exhaustive experiments involving different datasets and deep learning architectures, including fusion strategies and ensemble methods, we found that models trained on diverse datasets can achieve 95\% accuracy without the burden of data annotation and model training on the target parking lot

Copulas are a powerful statistical tool that captures dependencies across data dimensions. When applying Copulas, we can estimate multivariate distribution functions by initially estimating independent marginals, an easy task, and then a single copulating function, $C$, to connect the marginals, a hard task. For two-dimensional data, a copula is a two-increasing function of the form $C: (u,v)\in \mathbf{I}^2 \rightarrow \mathbf{I}$, where $\mathbf{I} = [0, 1]$. In this paper, we show how Neural Networks (NNs) can approximate any two-dimensional copula non-parametrically. Our approach, denoted as 2-Cats, is inspired by the Physics-Informed Neural Networks and Sobolev Training literature. Not only do we show that we can estimate the output of a 2d Copula better than the state-of-the-art, our approach is non-parametric and respects the mathematical properties of a Copula $C$.

Currently, low-resolution image recognition is confronted with a significant challenge in the field of intelligent traffic perception. Compared to high-resolution images, low-resolution images suffer from small size, low quality, and lack of detail, leading to a notable decrease in the accuracy of traditional neural network recognition algorithms. The key to low-resolution image recognition lies in effective feature extraction. Therefore, this paper delves into the fundamental dimensions of residual modules and their impact on feature extraction and computational efficiency. Based on experiments, we introduce a dual-branch residual network structure that leverages the basic architecture of residual networks and a common feature subspace algorithm. Additionally, it incorporates the utilization of intermediate-layer features to enhance the accuracy of low-resolution image recognition. Furthermore, we employ knowledge distillation to reduce network parameters and computational overhead. Experimental results validate the effectiveness of this algorithm for low-resolution image recognition in traffic environments.

We introduce Transformer-VQ, a decoder-only transformer computing softmaxbased dense self-attention in linear time. Transformer-VQ's efficient attention is enabled by vector-quantized keys and a novel caching mechanism. In large-scale experiments, Transformer-VQ is shown highly competitive in quality, with strong results on Enwik8 (0.99 bpb), PG-19 (26.6 ppl), and ImageNet64 (3.16 bpb). Figure 1: Minibatch of generated samples from our unconditional ImageNet64 model; nucleus 1.0. Transformer (Vaswani et al., 2017) language models would ideally scale to long sequences, since their predictive abilities often improve as context length increases (Dai et al., 2019; Kaplan et al., 2020). Unfortunately, the standard transformer uses a self-attention mechanism with a quadratic time complexity with respect to sequence length. Up to this point, a variety of efficient transformers (Tay et al., 2020b) have been proposed to scale to long sequences. Other efficient sequence models have also been proposed (Gu et al., 2022; Lee-Thorp et al., 2022; Poli et al., 2023; Peng et al., 2023).

Background: Atrial fibrillation (AF) is one of the most common cardiac arrhythmias that affects millions of people each year worldwide and it is closely linked to increased risk of cardiovascular diseases such as stroke and heart failure. Machine learning methods have shown promising results in evaluating the risk of developing atrial fibrillation from the electrocardiogram. We aim to develop and evaluate one such algorithm on a large CODE dataset collected in Brazil. Results: The deep neural network model identified patients without indication of AF in the presented ECG but who will develop AF in the future with an AUC score of 0.845. From our survival model, we obtain that patients in the high-risk group (i.e. with the probability of a future AF case being greater than 0.7) are 50% more likely to develop AF within 40 weeks, while patients belonging to the minimal-risk group (i.e. with the probability of a future AF case being less than or equal to 0.1) have more than 85% chance of remaining AF free up until after seven years. Conclusion: We developed and validated a model for AF risk prediction. If applied in clinical practice, the model possesses the potential of providing valuable and useful information in decision-making and patient management processes.