How to boost speech pre-training with textual data is an unsolved problem due to the fact that speech and text are very different modalities with distinct characteristics. In this paper, we propose a cross-modal Speech and Language Model (SpeechLM) to explicitly align speech and text pre-training with a pre-defined unified discrete representation. Specifically, we introduce two alternative discrete tokenizers to bridge the speech and text modalities, including phoneme-unit and hidden-unit tokenizers, which can be trained using a small amount of paired speech-text data. Based on the trained tokenizers, we convert the unlabeled speech and text data into tokens of phoneme units or hidden units. The pre-training objective is designed to unify the speech and the text into the same discrete semantic space with a unified Transformer network. We evaluate SpeechLM on various spoken language processing tasks including speech recognition, speech translation, and universal representation evaluation framework SUPERB, demonstrating significant improvements on content-related tasks. Code and models are available at https://aka.ms/SpeechLM.

Despite the prevalence of wireless connectivity in urban areas around the globe, there remain numerous and diverse situations where connectivity is insufficient or unavailable. To address this, we introduce mobile wireless infrastructure on demand, a system of UAVs that can be rapidly deployed to establish an ad-hoc wireless network. This network has the capability of reconfiguring itself dynamically to satisfy and maintain the required quality of communication. The system optimizes the positions of the UAVs and the routing of data flows throughout the network to achieve this quality of service (QoS). By these means, task agents using the network simply request a desired QoS, and the system adapts accordingly while allowing them to move freely. We have validated this system both in simulation and in real-world experiments. The results demonstrate that our system effectively offers mobile wireless infrastructure on demand, extending the operational range of task agents and supporting complex mobility patterns, all while ensuring connectivity and being resilient to agent failures.

Speaker-independent VSR is a complex task that involves identifying spoken words or phrases from video recordings of a speaker's facial movements. Over the years, there has been a considerable amount of research in the field of VSR involving different algorithms and datasets to evaluate system performance. These efforts have resulted in significant progress in developing effective VSR models, creating new opportunities for further research in this area. This survey provides a detailed examination of the progression of VSR over the past three decades, with a particular emphasis on the transition from speaker-dependent to speaker-independent systems. We also provide a comprehensive overview of the various datasets used in VSR research and the preprocessing techniques employed to achieve speaker independence. The survey covers the works published from 1990 to 2023, thoroughly analyzing each work and comparing them on various parameters. This survey provides an in-depth analysis of speaker-independent VSR systems evolution from 1990 to 2023. It outlines the development of VSR systems over time and highlights the need to develop end-to-end pipelines for speaker-independent VSR. The pictorial representation offers a clear and concise overview of the techniques used in speaker-independent VSR, thereby aiding in the comprehension and analysis of the various methodologies. The survey also highlights the strengths and limitations of each technique and provides insights into developing novel approaches for analyzing visual speech cues. Overall, This comprehensive review provides insights into the current state-of-the-art speaker-independent VSR and highlights potential areas for future research.

Text-to-image generation has seen an explosion of interest since 2021. Today, beautiful and intriguing digital images and artworks can be synthesized from textual inputs ("prompts") with deep generative models. Online communities around text-to-image generation and AI generated art have quickly emerged. This paper identifies six types of prompt modifiers used by practitioners in the online community based on a 3-month ethnographic study. The novel taxonomy of prompt modifiers provides researchers a conceptual starting point for investigating the practice of text-to-image generation, but may also help practitioners of AI generated art improve their images. We further outline how prompt modifiers are applied in the practice of "prompt engineering." We discuss research opportunities of this novel creative practice in the field of Human-Computer Interaction (HCI). The paper concludes with a discussion of broader implications of prompt engineering from the perspective of Human-AI Interaction (HAI) in future applications beyond the use case of text-to-image generation and AI generated art.

The diagnosis of faults present in a REM is integrated by the detection, identification and isolation of an anomaly, which can be achieved by using the information obtained on the state of operation of the equipment or drive [3]. As a result, it is possible to consider fault diagnosis as a pattern recognition problem with respect to the condition of a REM [4]. To effectively diagnose faults in a REM, it is essential to distinguish between failures originating from the machine itself, whether electrical or mechanical, and those corresponding to the associated load [5]. In recent decades, with the advancement of communication technologies and the inclusion of control devices in REM, non-invasive faults detection and diagnosis techniques based on the use of electrical variables have been studied more than those that use acoustic emissions, analysis lubrication, thermography and vibrations. The latter have been the techniques most widely used for some time, in which different methods are used for analysis, among the most common, Fast Fourier Transform (FFT) in the frequency domain, and wavelet analysis and empirical model decomposition in the domain time-frequency [6].

A major barrier to deploying healthcare AI models is their trustworthiness. One form of trustworthiness is a model's robustness across different subgroups: while existing models may exhibit expert-level performance on aggregate metrics, they often rely on non-causal features, leading to errors in hidden subgroups. To take a step closer towards trustworthy seizure onset detection from EEG, we propose to leverage annotations that are produced by healthcare personnel in routine clinical workflows -- which we refer to as workflow notes -- that include multiple event descriptions beyond seizures. Using workflow notes, we first show that by scaling training data to an unprecedented level of 68,920 EEG hours, seizure onset detection performance significantly improves (+12.3 AUROC points) compared to relying on smaller training sets with expensive manual gold-standard labels. Second, we reveal that our binary seizure onset detection model underperforms on clinically relevant subgroups (e.g., up to a margin of 6.5 AUROC points between pediatrics and adults), while having significantly higher false positives on EEG clips showing non-epileptiform abnormalities compared to any EEG clip (+19 FPR points). To improve model robustness to hidden subgroups, we train a multilabel model that classifies 26 attributes other than seizures, such as spikes, slowing, and movement artifacts. We find that our multilabel model significantly improves overall seizure onset detection performance (+5.9 AUROC points) while greatly improving performance among subgroups (up to +8.3 AUROC points), and decreases false positives on non-epileptiform abnormalities by 8 FPR points. Finally, we propose a clinical utility metric based on false positives per 24 EEG hours and find that our multilabel model improves this clinical utility metric by a factor of 2x across different clinical settings.

There are numerous advantages of deep neural network surrogate modeling for response time-history prediction. However, due to the high cost of refined numerical simulations and actual experiments, the lack of data has become an unavoidable bottleneck in practical applications. An iterative self-transfer learning method for training neural networks based on small datasets is proposed in this study. A new mapping-based transfer learning network, named as deep adaptation network with three branches for regression (DAN-TR), is proposed. A general iterative network training strategy is developed by coupling DAN-TR and the pseudo-label (PL) strategy, and the establishment of corresponding datasets is also discussed. Finally, a complex component is selected as a case study. The results show that the proposed method can improve the model performance by near an order of magnitude on small datasets without the need of external labeled samples, well behaved pre-trained models, additional artificial labeling, and complex physical/mathematical analysis.

The ability to connect language units to their referents in the physical world, referred to as grounding, is crucial to learning and understanding grounded meanings of words. While humans demonstrate fast mapping in new word learning, it remains unclear whether modern vision-language models can truly represent language with their grounded meanings and how grounding may further bootstrap new word learning. To this end, we introduce Grounded Open Vocabulary Acquisition (GOVA) to examine grounding and bootstrapping in open-world language learning. As an initial attempt, we propose object-oriented BERT (OctoBERT), a novel visually-grounded language model by pre-training on image-text pairs highlighting grounding as an objective. Through extensive experiments and analysis, we demonstrate that OctoBERT is a more coherent and fast grounded word learner, and that the grounding ability acquired during pre-training helps the model to learn unseen words more rapidly and robustly. Our code is available at https://github.com/sled-group/world-to-words

Vision-and-language (VL) models with separate encoders for each modality (e.g., CLIP) have become the go-to models for zero-shot image classification and image-text retrieval. The bulk of the evaluation of these models is, however, performed with English text only: the costly creation of language-specific image-caption datasets has limited multilingual VL benchmarks to a handful of high-resource languages. In this work, we introduce Babel-ImageNet, a massively multilingual benchmark that offers (partial) translations of 1000 ImageNet labels to 92 languages, built without resorting to machine translation (MT) or requiring manual annotation. We instead automatically obtain reliable translations of ImageNext concepts by linking them -- via shared WordNet synsets -- to BabelNet, a massively multilingual lexico-semantic network. We evaluate 8 different publicly available multilingual CLIP models on zero-shot image classification (ZS-IC) for each of the 92 Babel-ImageNet languages, demonstrating a significant gap between English ImageNet performance and that of high-resource languages (e.g., German or Chinese), and an even bigger gap for low-resource languages (e.g., Sinhala or Lao). Crucially, we show that the models' ZS-IC performance on Babel-ImageNet highly correlates with their performance in image-text retrieval, validating that Babel-ImageNet is suitable for estimating the quality of the multilingual VL representation spaces for the vast majority of languages that lack gold image-text data. Finally, we show that the performance of multilingual CLIP for low-resource languages can be drastically improved via cheap, parameter-efficient language-specific training. We make our code and data publicly available: \url{https://github.com/gregor-ge/Babel-ImageNet}

We introduce anticipation: a method for constructing a controllable generative model of a temporal point process (the event process) conditioned asynchronously on realizations of a second, correlated process (the control process). We achieve this by interleaving sequences of events and controls, such that controls appear following stopping times in the event sequence. This work is motivated by problems arising in the control of symbolic music generation. We focus on infilling control tasks, whereby the controls are a subset of the events themselves, and conditional generation completes a sequence of events given the fixed control events. We train anticipatory infilling models using the large and diverse Lakh MIDI music dataset. These models match the performance of autoregressive models for prompted music generation, with the additional capability to perform infilling control tasks, including accompaniment. Human evaluators report that an anticipatory model produces accompaniments with similar musicality to even music composed by humans over a 20-second clip.