Forced alignment (FA) predicts start and end timestamps for words or characters in speech, but existing methods are language-specific and prone to cumulative temporal shifts.The multilingual speech understanding and longsequence processing abilities of speech large language models (SLLMs) make them promising for FA in multilingual, crosslingual, and long-form speech settings.However, directly applying the next-token prediction paradigm of SLLMs to FA results in hallucinations and slow inference.To bridge the gap, we propose LLM-ForcedAligner, reformulating FA as a slot-filling paradigm: timestamps are treated as discrete indices, and special timestamp tokens are inserted as slots into the transcript.Conditioned on the speech embeddings and the transcript with slots, the SLLM directly predicts the time indices at slots.During training, causal attention masking with non-shifted input and label sequences allows each slot to predict its own timestamp index based on itself and preceding context, with loss computed only at slot positions.Dynamic slot insertion enables FA at arbitrary positions.Moreover, non-autoregressive inference is supported, avoiding hallucinations and improving speed.Experiments across multilingual, crosslingual, and long-form speech scenarios show that LLM-ForcedAligner achieves a 69%~78% relative reduction in accumulated averaging shift compared with prior methods.Checkpoint and inference code are available at https://huggingface.co/Qwen/ Qwen3-ForcedAligner-0.6B.

Building upon advancements in Large Language Models (LLMs), the field of audio processing has seen increased interest in training speech generation tasks with discrete speech token sequences.However, directly discretizing speech by neural audio codecs often results in sequences that fundamentally differ from text sequences.Unlike text, where text token sequences are deterministic, discrete speech tokens can exhibit significant variability based on contextual factors, while still producing perceptually identical audio segments.We refer to this phenomenon as Discrete Representation Inconsistency (DRI).This inconsistency can lead to a single speech segment being represented by multiple divergent sequences, which creates confusion in neural codec language models and results in poor generated speech.In this paper, we quantitatively analyze the DRI phenomenon within popular audio tokenizers such as En-Codec.Our approach effectively mitigates the DRI phenomenon of the neural audio codec.Furthermore, extensive experiments on the neural codec language model over LibriTTS and large-scale MLS dataset (44,000 hours) demonstrate the effectiveness and generality of our method.The demo of audio samples is available online 1 .

Neural audio codecs, used as speech tokenizers, have demonstrated remarkable potential in the field of speech generation. However, to ensure high-fidelity audio reconstruction, neural audio codecs typically encode audio into long sequences of speech tokens, posing a significant challenge for downstream language models in long-context modeling. We observe that speech token sequences exhibit short-range dependency: due to the monotonic alignment between text and speech in text-to-speech (TTS) tasks, the prediction of the current token primarily relies on its local context, while long-range tokens contribute less to the current token prediction and often contain redundant information. Inspired by this observation, we propose a compressed-to-fine language modeling approach to address the challenge of long sequence speech tokens within neural codec language models: (1) Fine-grained Initial and Short-range Information: Our approach retains the prompt and local tokens during prediction to ensure text alignment and the integrity of paralinguistic information; (2) Compressed Long-range Context: Our approach compresses long-range token spans into compact representations to reduce redundant information while preserving essential semantics. Extensive experiments on various neural audio codecs and downstream language models validate the effectiveness and generalizability of the proposed approach, highlighting the importance of token compression in improving speech generation within neural codec language models. The demo of audio samples will be available at https://anonymous.4open.science/r/SpeechTokenPredictionViaCompressedToFinedLM.

Qian Yang, Jin Xu, Wenrui Liu, Yunfei Chu, Ziyue Jiang, Xiaohuan Zhou, Yichong Leng, Yuanjun Lv, Zhou Zhao, Chang Zhou, Jingren Zhou. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2024.

Graph Neural Networks (GNNs) have already been widely used in various graph mining tasks. However, recent works reveal that the learned weights (channels) in well-trained GNNs are highly redundant, which inevitably limits the performance of GNNs. Instead of removing these redundant channels for efficiency consideration, we aim to reactivate them to enlarge the representation capacity of GNNs for effective graph learning. In this paper, we propose to substitute these redundant channels with other informative channels to achieve this goal. We introduce a novel GNN learning framework named AKE-GNN, which performs the Adaptive Knowledge Exchange strategy among multiple graph views generated by graph augmentations. AKE-GNN first trains multiple GNNs each corresponding to one graph view to obtain informative channels. Then, AKE-GNN iteratively exchanges redundant channels in the weight parameter matrix of one GNN with informative channels of another GNN in a layer-wise manner. Additionally, existing GNNs can be seamlessly incorporated into our framework. AKE-GNN achieves superior performance compared with various baselines across a suite of experiments on node classification, link prediction, and graph classification. In particular, we conduct a series of experiments on 15 public benchmark datasets, 8 popular GNN models, and 3 graph tasks and show that AKE-GNN consistently outperforms existing popular GNN models and even their ensembles. Extensive ablation studies and analyses on knowledge exchange methods validate the effectiveness of AKE-GNN.

Graph Neural Networks (GNNs) have become increasingly popular and achieved impressive results in many graph-based applications. However, extensive manual work and domain knowledge are required to design effective architectures, and the results of GNN models have high variance with different training setups, which limits the application of existing GNN models. In this paper, we present AutoHEnsGNN, a framework to build effective and robust models for graph tasks without any human intervention. AutoHEnsGNN won first place in the AutoGraph Challenge for KDD Cup 2020, and achieved the best rank score of five real-life datasets in the final phase. Given a task, AutoHEnsGNN first applies a fast proxy evaluation to automatically select a pool of promising GNN models. Then it builds a hierarchical ensemble framework: 1) We propose graph self-ensemble (GSE), which can reduce the variance of weight initialization and efficiently exploit the information of local and global neighborhoods; 2) Based on GSE, a weighted ensemble of different types of GNN models is used to effectively learn more discriminative node representations. To efficiently search the architectures and ensemble weights, we propose AutoHEnsGNNGradient, which treats the architectures and ensemble weights as architecture parameters and uses gradient-based architecture search to obtain optimal configurations, and AutoHEnsGNNAdaptive, which can adaptively adjust the ensemble weight based on the model accuracy. Extensive experiments on node classification, graph classification, edge prediction and KDD Cup challenge demonstrate the effectiveness and generality of AutoHEnsGNN.

Weight sharing is a popular approach to reduce the training cost of neural architecture search (NAS) by reusing the weights of shared operators from previously trained child models. However, the rank correlation between the estimated accuracy and ground truth accuracy of those child models is low due to the interference among different child models caused by weight sharing. In this paper, we investigate the interference issue by sampling different child models and calculating the gradient similarity of shared operators, and observe that: 1) the interference on a shared operator between two child models is positively correlated with the number of different operators between them; 2) the interference is smaller when the inputs and outputs of the shared operator are more similar. Inspired by these two observations, we propose two approaches to mitigate the interference: 1) rather than randomly sampling child models for optimization, we propose a gradual modification scheme by modifying one operator between adjacent optimization steps to minimize the interference on the shared operators; 2) forcing the inputs and outputs of the operator across all child models to be similar to reduce the interference. Experiments on a BERT search space verify that mitigating interference via each of our proposed methods improves the rank correlation of super-pet and combining both methods can achieve better results. Our discovered architecture outperforms RoBERTa$_{\rm base}$ by 1.1 and 0.6 points and ELECTRA$_{\rm base}$ by 1.6 and 1.1 points on the dev and test set of GLUE benchmark. Extensive results on the BERT compression, reading comprehension and large-scale image classification tasks also demonstrate the effectiveness and generality of our proposed methods.

While large-scale neural language models, such as GPT2 and BART,have achieved impressive results on various text generation tasks, they tend to get stuck in undesirable sentence-level loops with maximization-based decoding algorithms (\textit{e.g.}, greedy search). This phenomenon is counter-intuitive since there are few consecutive sentence-level repetitions in the human corpus (e.g., 0.02\% in Wikitext-103). To investigate the underlying reasons for generating consecutive sentence-level repetitions, we study the relationship between the probability of repetitive tokens and their previous repetitions in context. Through our quantitative experiments, we find that 1) Models have a preference to repeat the previous sentence; 2) The sentence-level repetitions have a \textit{self-reinforcement effect}: the more times a sentence is repeated in the context, the higher the probability of continuing to generate that sentence; 3) The sentences with higher initial probabilities usually have a stronger self-reinforcement effect. Motivated by our findings, we propose a simple and effective training method \textbf{DITTO} (Pseu\underline{D}o-Repet\underline{IT}ion Penaliza\underline{T}i\underline{O}n), where the model learns to penalize probabilities of sentence-level repetitions from synthetic repetitive data. Although our method is motivated by mitigating repetitions, our experiments show that DITTO not only mitigates the repetition issue without sacrificing perplexity, but also achieves better generation quality. Extensive experiments on open-ended text generation (Wikitext-103) and text summarization (CNN/DailyMail) demonstrate the generality and effectiveness of our method.

In this paper, we propose MixSpeech, a simple yet effective data augmentation method based on mixup for automatic speech recognition (ASR). MixSpeech trains an ASR model by taking a weighted combination of two different speech features (e.g., mel-spectrograms or MFCC) as the input, and recognizing both text sequences, where the two recognition losses use the same combination weight. We apply MixSpeech on two popular end-to-end speech recognition models including LAS (Listen, Attend and Spell) and Transformer, and conduct experiments on several low-resource datasets including TIMIT, WSJ, and HKUST. Experimental results show that MixSpeech achieves better accuracy than the baseline models without data augmentation, and outperforms a strong data augmentation method SpecAugment on these recognition tasks. Specifically, MixSpeech outperforms SpecAugment with a relative PER improvement of 10.6% on TIMIT dataset, and achieves a strong WER of 4.7% on WSJ dataset.

While pre-trained language models (e.g., BERT) have achieved impressive results on different natural language processing tasks, they have large numbers of parameters and suffer from big computational and memory costs, which make them difficult for real-world deployment. Therefore, model compression is necessary to reduce the computation and memory cost of pre-trained models. In this work, we aim to compress BERT and address the following two challenging practical issues: (1) The compression algorithm should be able to output multiple compressed models with different sizes and latencies, in order to support devices with different memory and latency limitations; (2) The algorithm should be downstream task agnostic, so that the compressed models are generally applicable for different downstream tasks. We leverage techniques in neural architecture search (NAS) and propose NAS-BERT, an efficient method for BERT compression. NAS-BERT trains a big supernet on a carefully designed search space containing a variety of architectures and outputs multiple compressed models with adaptive sizes and latency. Furthermore, the training of NAS-BERT is conducted on standard self-supervised pre-training tasks (e.g., masked language model) and does not depend on specific downstream tasks. Thus, the compressed models can be used across various downstream tasks. The technical challenge of NAS-BERT is that training a big supernet on the pre-training task is extremely costly. We employ several techniques including block-wise search, search space pruning, and performance approximation to improve search efficiency and accuracy. Extensive experiments on GLUE and SQuAD benchmark datasets demonstrate that NAS-BERT can find lightweight models with better accuracy than previous approaches, and can be directly applied to different downstream tasks with adaptive model sizes for different requirements of memory or latency.

Neural Transducer (e.g., RNN-T) has been widely used in automatic speech recognition (ASR) due to its capabilities of efficiently modeling monotonic alignments between input and output sequences and naturally supporting streaming inputs. Considering that monotonic alignments are also critical to text to speech (TTS) synthesis and streaming TTS is also an important application scenario, in this work, we explore the possibility of applying Transducer to TTS and more. However, it is challenging because it is difficult to trade off the emission (continuous mel-spectrogram prediction) probability and transition (ASR Transducer predicts blank token to indicate transition to next input) probability when calculating the output probability lattice in Transducer, and it is not easy to learn the alignments between text and speech through the output probability lattice. We propose SpeechTransducer (Speech-T for short), a Transformer based Transducer model that 1) uses a new forward algorithm to separate the transition prediction from the continuous mel-spectrogram prediction when calculating the output probability lattice, and uses a diagonal constraint in the probability lattice to help the alignment learning; 2) supports both full-sentence or streaming TTS by adjusting the look-ahead context; and 3) further supports both TTS and ASR together for the first time, which enjoys several advantages including fewer parameters as well as streaming synthesis and recognition in a single model. Experiments on LJSpeech datasets demonstrate that Speech-T 1) is more robust than the attention based autoregressive TTS model due to its inherent monotonic alignments between text and speech; 2) naturally supports streaming TTS with good voice quality; and 3) enjoys the benefit of joint modeling TTS and ASR in a single network.

Error correction techniques have been used to refine the output sentences from automatic speech recognition (ASR) models and achieve a lower word error rate (WER) than original ASR outputs. Previous works usually use a sequence-to-sequence model to correct an ASR output sentence autoregressively, which causes large latency and cannot be deployed in online ASR services. A straightforward solution to reduce latency, inspired by non-autoregressive (NAR) neural machine translation, is to use an NAR sequence generation model for ASR error correction, which, however, comes at the cost of significantly increased ASR error rate. In this paper, observing distinctive error patterns and correction operations (i.e., insertion, deletion, and substitution) in ASR, we propose FastCorrect, a novel NAR error correction model based on edit alignment. In training, FastCorrect aligns each source token from an ASR output sentence to the target tokens from the corresponding ground-truth sentence based on the edit distance between the source and target sentences, and extracts the number of target tokens corresponding to each source token during edition/correction, which is then used to train a length predictor and to adjust the source tokens to match the length of the target sentence for parallel generation. In inference, the token number predicted by the length predictor is used to adjust the source tokens for target sequence generation. Experiments on the public AISHELL-1 dataset and an internal industrial-scale ASR dataset show the effectiveness of FastCorrect for ASR error correction: 1) it speeds up the inference by 6-9 times and maintains the accuracy (8-14% WER reduction) compared with the autoregressive correction model; and 2) it outperforms the popular NAR models adopted in neural machine translation and text edition by a large margin.

Revenue prediction is an essential component in security analysis since the revenue of a company has a great impact on the performance of its stock. For investment, one of the most valuable pieces of information is the company’s unexpected revenue, which is the difference between the officially reported revenue and the consensus estimate for revenue predicted by analysts. Since it is the unexpected revenue that indicates something exceeding or under analysts’ expectation, it is an indispensable factor that influences the performance of a stock. Besides conventional trading data from stock market and companies’ financial reports, recent years have witnessed an extensive application of alternative data for gaining an information edge in stock investment.In this paper, we study the challenging problem of better predicting unexpected revenue of a company via machine learning with alternative data. To the best of our knowledge, this is the first work studying this problem in literature. However, it is nontrivial to quantitatively model the relations between the unexpected revenue and the information provided by alternative data with a machine learning approach. Thus we proposed an adaptive master-slave regularized model, called AMS for short, to effectively leverage alternative data for unexpected revenue prediction. AMS first trains a master model upon a company graph, which captures the relations among companies, using a graph neural network (GNN). Then for a target company, the master model generates an adaptive slave-model, which is specially optimized for this target company. Finally, we use this slave-model to predict the unexpected revenue of the target company. Besides its excellent prediction performance, another critical advantage of our AMS model lies in its superior interpretability, which is crucial for portfolio managers to understand the predicted results. With extensive experiments using two real-world alternative datasets, we have demonstrated the effectiveness of our model against a set of competitors.

Transformer-based text to speech (TTS) model (e.g., Transformer TTS [1], FastSpeech [2]) has shown the advantages of training and inference efficiency over Tacotron [3]) due to its parallel computation in training and/or inference.However, the parallel computation increases the difficulty while learning the alignment between text and speech in Transformer, which is further magnified in the multi-speaker scenario with noisy data and diverse speakers, and hinders the applicability of Transformer for multi-speaker TTS.In this paper, we develop a robust and high-quality multispeaker Transformer TTS system called MultiSpeech, with several specially designed components/techniques to improve textto-speech alignment: 1) a diagonal constraint on the weight matrix of encoder-decoder attention in both training and inference; 2) layer normalization on phoneme embedding in encoder to better preserve position information; 3) a bottleneck in decoder pre-net to prevent copy between consecutive speech frames.Experiments on VCTK and LibriTTS multi-speaker datasets demonstrate the effectiveness of MultiSpeech: 1) it synthesizes more robust and better quality multi-speaker voice than naive Transformer based TTS; 2) with a MutiSpeech model as the teacher, we obtain a strong multi-speaker FastSpeech model with almost zero quality degradation while enjoying extremely fast inference speed.

Speech synthesis (text to speech, TTS) and recognition (automatic speech recognition, ASR) are important speech tasks, and require a large amount of text and speech pairs for model training. However, there are more than 6,000 languages in the world and most languages are lack of speech training data, which poses significant challenges when building TTS and ASR systems for extremely low-resource languages. In this paper, we develop LRSpeech, a TTS and ASR system under the extremely low-resource setting, which can support rare languages with low data cost. LRSpeech consists of three key techniques: 1) pre-training on rich-resource languages and fine-tuning on low-resource languages; 2) dual transformation between TTS and ASR to iteratively boost the accuracy of each other; 3) knowledge distillation to customize the TTS model on a high-quality target-speaker voice and improve the ASR model on multiple voices. We conduct experiments on an experimental language (English) and a truly low-resource language (Lithuanian) to verify the effectiveness of LRSpeech. Experimental results show that LRSpeech 1) achieves high quality for TTS in terms of both intelligibility (more than $98%$ intelligibility rate) and naturalness (above 3.5 mean opinion score (MOS)) of the synthesized speech, which satisfy the requirements for industrial deployment, 2) achieves promising recognition accuracy for ASR, and 3) last but not least, uses extremely low-resource training data. We also conduct comprehensive analyses on LRSpeech with different amounts of data resources, and provide valuable insights and guidances for industrial deployment. We are currently deploying LRSpeech into a commercialized cloud speech service to support TTS on more rare languages.

Tags of a Point of Interest (POI) can facilitate location-based services from many aspects like location search and place recommendation. However, many POI tags are often incomplete or imprecise, which may lead to performance degradation of tag-dependent applications. In this paper, we study the POI tag refinement problem which aims to automatically fill in the missing tags as well as correct noisy tags for POIs. We propose a tri-adaptive collaborative learning framework to search for an optimal POI-tag score matrix. The framework integrates three components to collaboratively (i) model the similarity matching between POI and tag, (ii) recover the POI-tag pattern via matrix factorization and (iii) learn to infer the most possible tags by maximum likelihood estimation. We devise an adaptively joint training process to optimize the model and regularize each component simultaneously. And the final refinement results are the consensus of multiple views from different components. We also discuss how to utilize various data sources to construct features for tag refinement, including user profile data, query data on Baidu Maps and basic properties of POIs. Finally, we conduct extensive experiments to demonstrate the effectiveness of our framework. And we further present a case study of the deployment of our framework on Baidu Maps.