In this work, we propose an error correction framework, named DiaCorrect, to refine the output of a diarization system in a simple yet effective way. This method is inspired by error correction techniques in automatic speech recognition. Our model consists of two parallel convolutional encoders and a transform-based decoder. By exploiting the interactions between the input recording and the initial system's outputs, DiaCorrect can automatically correct the initial speaker activities to minimize the diarization errors. Experiments on 2-speaker telephony data show that the proposed DiaCorrect can effectively improve the initial model's results. Our source code is publicly available at https://github.com/BUTSpeechFIT/diacorrect.
Modern approaches have proved the huge potential of addressing semantic segmentation as a mask classification task which is widely used in instance-level segmentation. This paradigm trains models by assigning part of object queries to ground truths via conventional one-to-one matching. However, we observe that the popular video semantic segmentation (VSS) dataset has limited categories per video, meaning less than 10% of queries could be matched to receive meaningful gradient updates during VSS training. This inefficiency limits the full expressive potential of all queries.Thus, we present a novel solution THE-Mask for VSS, which introduces temporal-aware hierarchical object queries for the first time. Specifically, we propose to use a simple two-round matching mechanism to involve more queries matched with minimal cost during training while without any extra cost during inference. To support our more-to-one assignment, in terms of the matching results, we further design a hierarchical loss to train queries with their corresponding hierarchy of primary or secondary. Moreover, to effectively capture temporal information across frames, we propose a temporal aggregation decoder that fits seamlessly into the mask-classification paradigm for VSS. Utilizing temporal-sensitive multi-level queries, our method achieves state-of-the-art performance on the latest challenging VSS benchmark VSPW without bells and whistles.
We propose a decoder-only language model, VoxtLM, that can perform four tasks: speech recognition, speech synthesis, text generation, and speech continuation. VoxtLM integrates text vocabulary with discrete speech tokens from self-supervised speech features and uses special tokens to enable multitask learning. Compared to a single-task model, VoxtLM exhibits a significant improvement in speech synthesis, with improvements in both speech intelligibility from 28.9 to 5.6 and objective quality from 2.68 to 3.90. VoxtLM also improves speech generation and speech recognition performance over the single-task counterpart. Further, VoxtLM is trained with publicly available data and training recipes and model checkpoints are open-sourced to make fully reproducible work.
Never having seen an object and heard its sound simultaneously, can the model still accurately localize its visual position from the input audio? In this work, we concentrate on the Audio-Visual Localization and Segmentation tasks but under the demanding zero-shot and few-shot scenarios. To achieve this goal, different from existing approaches that mostly employ the encoder-fusion-decoder paradigm to decode localization information from the fused audio-visual feature, we introduce the encoder-prompt-decoder paradigm, aiming to better fit the data scarcity and varying data distribution dilemmas with the help of abundant knowledge from pre-trained models. Specifically, we first propose to construct Semantic-aware Audio Prompt (SAP) to help the visual foundation model focus on sounding objects, meanwhile, the semantic gap between the visual and audio modalities is also encouraged to shrink. Then, we develop a Correlation Adapter (ColA) to keep minimal training efforts as well as maintain adequate knowledge of the visual foundation model. By equipping with these means, extensive experiments demonstrate that this new paradigm outperforms other fusion-based methods in both the unseen class and cross-dataset settings. We hope that our work can further promote the generalization study of Audio-Visual Localization and Segmentation in practical application scenarios.
Visual-based 3D semantic occupancy perception is a key technology for robotics, including autonomous vehicles, offering an enhanced understanding of the environment by 3D. This approach, however, typically requires more computational resources than BEV or 2D methods. We propose a novel 3D semantic occupancy perception method, OccupancyDETR, which utilizes a DETR-like object detection, a mixed dense-sparse 3D occupancy decoder. Our approach distinguishes between foreground and background within a scene. Initially, foreground objects are detected using the DETR-like object detection. Subsequently, queries for both foreground and background objects are fed into the mixed dense-sparse 3D occupancy decoder, performing upsampling in dense and sparse methods, respectively. Finally, a MaskFormer is utilized to infer the semantics of the background voxels. Our approach strikes a balance between efficiency and accuracy, achieving faster inference times, lower resource consumption, and improved performance for small object detection. We demonstrate the effectiveness of our proposed method on the SemanticKITTI dataset, showcasing an mIoU of 14 and a processing speed of 10 FPS, thereby presenting a promising solution for real-time 3D semantic occupancy perception.
Real-time transportation surveillance is an essential part of the intelligent transportation system (ITS). However, images captured under low-light conditions often suffer the poor visibility with types of degradation, such as noise interference and vague edge features, etc. With the development of imaging devices, the quality of the visual surveillance data is continually increasing, like 2K and 4K, which has more strict requirements on the efficiency of image processing. To satisfy the requirements on both enhancement quality and computational speed, this paper proposes a double domain guided real-time low-light image enhancement network (DDNet) for ultra-high-definition (UHD) transportation surveillance. Specifically, we design an encoder-decoder structure as the main architecture of the learning network. In particular, the enhancement processing is divided into two subtasks (i.e., color enhancement and gradient enhancement) via the proposed coarse enhancement module (CEM) and LoG-based gradient enhancement module (GEM), which are embedded in the encoder-decoder structure. It enables the network to enhance the color and edge features simultaneously. Through the decomposition and reconstruction on both color and gradient domains, our DDNet can restore the detailed feature information concealed by the darkness with better visual quality and efficiency. The evaluation experiments on standard and transportation-related datasets demonstrate that our DDNet provides superior enhancement quality and efficiency compared with the state-of-the-art methods. Besides, the object detection and scene segmentation experiments indicate the practical benefits for higher-level image analysis under low-light environments in ITS.
Video saliency prediction and detection are thriving research domains that enable computers to simulate the distribution of visual attention akin to how humans perceiving dynamic scenes. While many approaches have crafted task-specific training paradigms for either video saliency prediction or video salient object detection tasks, few attention has been devoted to devising a generalized saliency modeling framework that seamlessly bridges both these distinct tasks. In this study, we introduce the Unified Saliency Transformer (UniST) framework, which comprehensively utilizes the essential attributes of video saliency prediction and video salient object detection. In addition to extracting representations of frame sequences, a saliency-aware transformer is designed to learn the spatio-temporal representations at progressively increased resolutions, while incorporating effective cross-scale saliency information to produce a robust representation. Furthermore, a task-specific decoder is proposed to perform the final prediction for each task. To the best of our knowledge, this is the first work that explores designing a transformer structure for both saliency modeling tasks. Convincible experiments demonstrate that the proposed UniST achieves superior performance across seven challenging benchmarks for two tasks, and significantly outperforms the other state-of-the-art methods.
This paper works on non-autoregressive automatic speech recognition. A unimodal aggregation (UMA) is proposed to segment and integrate the feature frames that belong to the same text token, and thus to learn better feature representations for text tokens. The frame-wise features and weights are both derived from an encoder. Then, the feature frames with unimodal weights are integrated and further processed by a decoder. Connectionist temporal classification (CTC) loss is applied for training. Compared to the regular CTC, the proposed method learns better feature representations and shortens the sequence length, resulting in lower recognition error and computational complexity. Experiments on three Mandarin datasets show that UMA demonstrates superior or comparable performance to other advanced non-autoregressive methods, such as self-conditioned CTC. Moreover, by integrating self-conditioned CTC into the proposed framework, the performance can be further noticeably improved.
Automated Audio Captioning (AAC) is the task of generating natural language descriptions given an audio stream. A typical AAC system requires manually curated training data of audio segments and corresponding text caption annotations. The creation of these audio-caption pairs is costly, resulting in general data scarcity for the task. In this work, we address this major limitation and propose an approach to train AAC systems using only text. Our approach leverages the multimodal space of contrastively trained audio-text models, such as CLAP. During training, a decoder generates captions conditioned on the pretrained CLAP text encoder. During inference, the text encoder is replaced with the pretrained CLAP audio encoder. To bridge the modality gap between text and audio embeddings, we propose the use of noise injection or a learnable adapter, during training. We find that the proposed text-only framework performs competitively with state-of-the-art models trained with paired audio, showing that efficient text-to-audio transfer is possible. Finally, we showcase both stylized audio captioning and caption enrichment while training without audio or human-created text captions.
In deep learning, the load data with non-temporal factors are difficult to process by sequence models. This problem results in insufficient precision of the prediction. Therefore, a short-term load forecasting method based on convolutional neural network (CNN), self-attention encoder-decoder network (SAEDN) and residual-refinement (Res) is proposed. In this method, feature extraction module is composed of a two-dimensional convolutional neural network, which is used to mine the local correlation between data and obtain high-dimensional data features. The initial load fore-casting module consists of a self-attention encoder-decoder network and a feedforward neural network (FFN). The module utilizes self-attention mechanisms to encode high-dimensional features. This operation can obtain the global correlation between data. Therefore, the model is able to retain important information based on the coupling relationship between the data in data mixed with non-time series factors. Then, self-attention decoding is per-formed and the feedforward neural network is used to regression initial load. This paper introduces the residual mechanism to build the load optimization module. The module generates residual load values to optimize the initial load. The simulation results show that the proposed load forecasting method has advantages in terms of prediction accuracy and prediction stability.
Given a group of images, co-salient object detection (CoSOD) aims to highlight the common salient object in each image. There are two factors closely related to the success of this task, namely consensus extraction, and the dispersion of consensus to each image. Most previous works represent the group consensus using local features, while we instead utilize a hierarchical Transformer module for extracting semantic-level consensus. Therefore, it can obtain a more comprehensive representation of the common object category, and exclude interference from other objects that share local similarities with the target object. In addition, we propose a Transformer-based dispersion module that takes into account the variation of the co-salient object in different scenes. It distributes the consensus to the image feature maps in an image-specific way while making full use of interactions within the group. These two modules are integrated with a ViT encoder and an FPN-like decoder to form an end-to-end trainable network, without additional branch and auxiliary loss. The proposed method is evaluated on three commonly used CoSOD datasets and achieves state-of-the-art performance.
The attention-based encoder-decoder (AED) speech recognition model has been widely successful in recent years. However, the joint optimization of acoustic model and language model in end-to-end manner has created challenges for text adaptation. In particular, effective, quick and inexpensive adaptation with text input has become a primary concern for deploying AED systems in the industry. To address this issue, we propose a novel model, the hybrid attention-based encoder-decoder (HAED) speech recognition model that preserves the modularity of conventional hybrid automatic speech recognition systems. Our HAED model separates the acoustic and language models, allowing for the use of conventional text-based language model adaptation techniques. We demonstrate that the proposed HAED model yields 23% relative Word Error Rate (WER) improvements when out-of-domain text data is used for language model adaptation, with only a minor degradation in WER on a general test set compared with the conventional AED model.
This paper proposes a direct text to speech translation system using discrete acoustic units. This framework employs text in different source languages as input to generate speech in the target language without the need for text transcriptions in this language. Motivated by the success of acoustic units in previous works for direct speech to speech translation systems, we use the same pipeline to extract the acoustic units using a speech encoder combined with a clustering algorithm. Once units are obtained, an encoder-decoder architecture is trained to predict them. Then a vocoder generates speech from units. Our approach for direct text to speech translation was tested on the new CVSS corpus with two different text mBART models employed as initialisation. The systems presented report competitive performance for most of the language pairs evaluated. Besides, results show a remarkable improvement when initialising our proposed architecture with a model pre-trained with more languages.
Maximizing influences in complex networks is a practically important but computationally challenging task for social network analysis, due to its NP- hard nature. Most current approximation or heuristic methods either require tremendous human design efforts or achieve unsatisfying balances between effectiveness and efficiency. Recent machine learning attempts only focus on speed but lack performance enhancement. In this paper, different from previous attempts, we propose an effective deep reinforcement learning model that achieves superior performances over traditional best influence maximization algorithms. Specifically, we design an end-to-end learning framework that combines graph neural network as the encoder and reinforcement learning as the decoder, named DREIM. Trough extensive training on small synthetic graphs, DREIM outperforms the state-of-the-art baseline methods on very large synthetic and real-world networks on solution quality, and we also empirically show its linear scalability with regard to the network size, which demonstrates its superiority in solving this problem.
The task of blind source separation (BSS) involves separating sources from a mixture without prior knowledge of the sources or the mixing system. Single-channel mixtures and non-linear mixtures are a particularly challenging problem in BSS. In this paper, we propose a novel method for addressing BSS with single-channel non-linear mixtures by leveraging the natural feature subspace specialization ability of multi-encoder autoencoders. During the training phase, our method unmixes the input into the separate encoding spaces of the multi-encoder network and then remixes these representations within the decoder for a reconstruction of the input. Then to perform source inference, we introduce a novel encoding masking technique whereby masking out all but one of the encodings enables the decoder to estimate a source signal. To this end, we also introduce a sparse mixing loss that encourages sparse remixing of source encodings throughout the decoder and a so-called zero reconstruction loss on the decoder for coherent source estimations. To analyze and evaluate our method, we conduct experiments on a toy dataset, designed to demonstrate this property of feature subspace specialization, and with real-world biosignal recordings from a polysomnography sleep study for extracting respiration from electrocardiogram and photoplethysmography signals.
Diffusion tensor cardiac magnetic resonance (DT-CMR) is a method capable of providing non-invasive measurements of myocardial microstructure. Image registration is essential to correct image shifts due to intra and inter breath-hold motion and imperfect cardiac triggering. Registration is challenging in DT-CMR due to the low signal-to-noise and various contrasts induced by the diffusion encoding in the myocardium and surrounding organs. Traditional deformable registration corrects through-plane motion but at the risk of destroying the texture information while rigid registration inefficiently discards frames with local deformation. In this study, we explored the possibility of deep learning-based deformable registration on DT-CMR. Based on the noise suppression using low-rank features and diffusion encoding suppression using variational auto encoder-decoder, a B-spline based registration network extracted the displacement fields and maintained the texture features of DT-CMR. In this way, our method improved the efficiency of frame utilization, manual cropping, and computational speed.
To enhance the reasoning capabilities of off-the-shelf Large Language Models (LLMs), we introduce a simple, yet general and effective prompting method, Re2, i.e., \textbf{Re}-\textbf{Re}ading the question as input. Unlike most thought-eliciting prompting methods, such as Chain-of-Thought (CoT), which aim to elicit the reasoning process in the output, Re2 shifts the focus to the input by processing questions twice, thereby enhancing the understanding process. Consequently, Re2 demonstrates strong generality and compatibility with most thought-eliciting prompting methods, including CoT. Crucially, Re2 facilitates a "bidirectional" encoding in unidirectional decoder-only LLMs because the first pass could provide global information for the second pass. We begin with a preliminary empirical study as the foundation of Re2, illustrating its potential to enable "bidirectional" attention mechanisms. We then evaluate Re2 on extensive reasoning benchmarks across 14 datasets, spanning 112 experiments, to validate its effectiveness and generality. Our findings indicate that, with the exception of a few scenarios on vanilla ChatGPT, Re2 consistently enhances the reasoning performance of LLMs through a simple re-reading strategy. Further analyses reveal Re2's adaptability, showing how it can be effectively integrated with different LLMs, thought-eliciting prompting, and ensemble strategies. Our code is available at \url{https://github.com/Tebmer/Rereading-LLM-Reasoning/}
This paper proposes a new task in the field of Answering Subjective Induction Question on Products (SUBJPQA). The answer to this kind of question is non-unique, but can be interpreted from many perspectives. For example, the answer to 'whether the phone is heavy' has a variety of different viewpoints. A satisfied answer should be able to summarize these subjective opinions from multiple sources and provide objective knowledge, such as the weight of a phone. That is quite different from the traditional QA task, in which the answer to a factoid question is unique and can be found from a single data source. To address this new task, we propose a three-steps method. We first retrieve all answer-related clues from multiple knowledge sources on facts and opinions. The implicit commonsense facts are also collected to supplement the necessary but missing contexts. We then capture their relevance with the questions by interactive attention. Next, we design a reinforcement-based summarizer to aggregate all these knowledgeable clues. Based on a template-controlled decoder, we can output a comprehensive and multi-perspective answer. Due to the lack of a relevant evaluated benchmark set for the new task, we construct a large-scale dataset, named SupQA, consisting of 48,352 samples across 15 product domains. Evaluation results show the effectiveness of our approach.
Visual narrating focuses on generating semantic descriptions to summarize visual content of images or videos, e.g., visual captioning and visual storytelling. The challenge mainly lies in how to design a decoder to generate accurate descriptions matching visual content. Recent advances often employ a recurrent neural network (RNN), e.g., Long-Short Term Memory (LSTM), as the decoder. However, RNN is prone to diluting long-term information, which weakens its performance of capturing long-term dependencies. Recent work has demonstrated memory network (MemNet) owns the advantage of storing long-term information. However, as the decoder, it has not been well exploited for visual narrating. The reason partially comes from the difficulty of multi-modal sequential decoding with MemNet. In this article, we devise a novel memory decoder for visual narrating. Concretely, to obtain a better multi-modal representation, we first design a new multi-modal fusion method to fully merge visual and lexical information. Then, based on the fusion result, during decoding, we construct a MemNet-based decoder consisting of multiple memory layers. Particularly, in each layer, we employ a memory set to store previous decoding information and utilize an attention mechanism to adaptively select the information related to the current output. Meanwhile, we also employ a memory set to store the decoding output of each memory layer at the current time step and still utilize an attention mechanism to select the related information. Thus, this decoder alleviates dilution of long-term information. Meanwhile, the hierarchical architecture leverages the latent information of each layer, which is helpful for generating accurate descriptions. Experimental results on two tasks of visual narrating, i.e., video captioning and visual storytelling, show that our decoder could obtain superior results and outperform the performance of conventional RNN-based decoder.
Modulo sampling has recently drawn a great deal of attention for cutting-edge applications, due to overcoming the barrier of information loss through sensor saturation and clipping. This is a significant problem, especially when the range of signal amplitudes is unknown or in the near-far case. To overcome this fundamental bottleneck, we propose a one-bit-aided (1bit-aided) modulo sampling scheme for direction-of-arrival (DOA) estimation. On the one hand, one-bit quantization involving a simple comparator offers the advantages of low-cost and low-complexity implementation. On the other hand, one-bit quantization provides an estimate of the normalized covariance matrix of the unquantized measurements via the arcsin law. The estimate of the normalized covariance matrix is used to implement blind integer-forcing (BIF) decoder to unwrap the modulo samples to construct the covariance matrix, and subspace methods can be used to perform the DOA estimation. Our approach named as 1bit-aided-BIF addresses the near-far problem well and overcomes the intrinsic low dynamic range of one-bit quantization. Numerical experiments validate the excellent performance of the proposed algorithm.