A Long Short-Term Memory(LSTM) Recurrent Neural Network(RNN) has driven tremendous improvements on an acoustic model based on Gaussian Mixture Model(GMM). However, these models based on a hybrid method require a force...A Long Short-Term Memory(LSTM) Recurrent Neural Network(RNN) has driven tremendous improvements on an acoustic model based on Gaussian Mixture Model(GMM). However, these models based on a hybrid method require a forced aligned Hidden Markov Model(HMM) state sequence obtained from the GMM-based acoustic model. Therefore, it requires a long computation time for training both the GMM-based acoustic model and a deep learning-based acoustic model. In order to solve this problem, an acoustic model using CTC algorithm is proposed. CTC algorithm does not require the GMM-based acoustic model because it does not use the forced aligned HMM state sequence. However, previous works on a LSTM RNN-based acoustic model using CTC used a small-scale training corpus. In this paper, the LSTM RNN-based acoustic model using CTC is trained on a large-scale training corpus and its performance is evaluated. The implemented acoustic model has a performance of 6.18% and 15.01% in terms of Word Error Rate(WER) for clean speech and noisy speech, respectively. This is similar to a performance of the acoustic model based on the hybrid method.展开更多
为了提高光伏发电功率预测精度,提出了一种基于长短期时序数据融合的Transformer生成式预测模型:LSTformer,能准确有效地预测光伏发电功率。LSTformer创新性地提出了时序分析模块(time series analysis,TSA)、时序特征融合模块(time ser...为了提高光伏发电功率预测精度,提出了一种基于长短期时序数据融合的Transformer生成式预测模型:LSTformer,能准确有效地预测光伏发电功率。LSTformer创新性地提出了时序分析模块(time series analysis,TSA)、时序特征融合模块(time series feature fusion,TSFF)和多周期嵌入模块(cycleEmbed),利用数据融合解决难以提取多时间尺度时序特征问题。设计时间卷积前馈(time convolution feedforward,TCNforward)单元,在编解码的过程中进一步提取时序特征。利用某光伏电站实际历史发电数据,通过实验验证LSTformer模型在光伏发电功率预测领域得到最低的均方误差(mean squared error,MSE)、平均绝对误差(mean absolute error,MAE),并通过消融实验验证了各模块的有效性。展开更多
基金supported by the Ministry of Trade,Industry & Energy(MOTIE,Korea) under Industrial Technology Innovation Program (No.10063424,'development of distant speech recognition and multi-task dialog processing technologies for in-door conversational robots')
文摘A Long Short-Term Memory(LSTM) Recurrent Neural Network(RNN) has driven tremendous improvements on an acoustic model based on Gaussian Mixture Model(GMM). However, these models based on a hybrid method require a forced aligned Hidden Markov Model(HMM) state sequence obtained from the GMM-based acoustic model. Therefore, it requires a long computation time for training both the GMM-based acoustic model and a deep learning-based acoustic model. In order to solve this problem, an acoustic model using CTC algorithm is proposed. CTC algorithm does not require the GMM-based acoustic model because it does not use the forced aligned HMM state sequence. However, previous works on a LSTM RNN-based acoustic model using CTC used a small-scale training corpus. In this paper, the LSTM RNN-based acoustic model using CTC is trained on a large-scale training corpus and its performance is evaluated. The implemented acoustic model has a performance of 6.18% and 15.01% in terms of Word Error Rate(WER) for clean speech and noisy speech, respectively. This is similar to a performance of the acoustic model based on the hybrid method.
基金This study was jointly funded by the National Key R&D Program of China[grant number 2022YFC3004103]the National Natural Foundation of China[grant number 42275003]+2 种基金the Beijing Science and Technology Program[grant number Z221100005222012]the Beijing Meteorological Service Science and Technology Program[grant number BMBKJ202302004]the China Meteorological Administration Youth Innovation Team[grant number CMA2023QN10].