Align With Purpose: Optimize Desired Properties in CTC Models with a General Plug-and-Play Framework
Connectionist Temporal Classification (CTC) is a widely used criterion for training supervised sequence-to-sequence (seq2seq) models. It enables learning the relations between input and output sequences, termed alignments, by marginalizing over perfect alignments (that yield the ground truth), at the expense of imperfect alignments. This binary differentiation of perfect and imperfect alignments falls short of capturing other essential alignment properties that hold significance in other real-world applications. Here we propose $\textit{Align With Purpose}$, a $\textbf{general Plug-and-Play framework}$ for enhancing a desired property in models trained with the CTC criterion. We do that by complementing the CTC with an additional loss term that prioritizes alignments according to a desired property. Our method does not require any intervention in the CTC loss function, enables easy optimization of a variety of properties, and allows differentiation between both perfect and imperfect alignments. We apply our framework in the domain of Automatic Speech Recognition (ASR) and show its generality in terms of property selection, architectural choice, and scale of training dataset (up to 280,000 hours). To demonstrate the effectiveness of our framework, we apply it to two unrelated properties: emission time and word error rate (WER). For the former, we report an improvement of up to 570ms in latency optimization with a minor reduction in WER, and for the latter, we report a relative improvement of 4.5% WER over the baseline models. To the best of our knowledge, these applications have never been demonstrated to work on a scale of data as large as ours. Notably, our method can be implemented using only a few lines of code, and can be extended to other alignment-free loss functions and to domains other than ASR.
Code (0)
등록된 구현이 없습니다.
Tasks
Automatic Speech RecognitionAutomatic Speech Recognition (ASR)speech-recognitionSpeech RecognitionMethods 이 논문이 사용한 방법론
Similar Papers 제목 키워드 기반
MGCVAE: Multi-objective Inverse Design via Molecular Graph Conditional Variational Autoencoder
The ultimate goal of various fields is to directly generate molecules with desired properties, such as finding water-soluble molecules in drug development and finding molecules suitable for organic light-emitting diode (…
Drug DiscoveryGeometrical analysis of polynomial lens distortion models
Polynomial functions are a usual choice to model the nonlinearity of lenses. Typically, these models are obtained through physical analysis of the lens system or on purely empirical grounds. The aim of this work is to fa…
OptiGAN: Generative Adversarial Networks for Goal Optimized Sequence Generation
One of the challenging problems in sequence generation tasks is the optimized generation of sequences with specific desired goals. Current sequential generative models mainly generate sequences to closely mimic the train…
Diversityreinforcement-learningReinforcement LearningReinforcement Learning (RL)Iterative Foundation Model Fine-Tuning on Multiple Rewards
Fine-tuning foundation models has emerged as a powerful approach for generating objects with specific desired properties. Reinforcement learning (RL) provides an effective framework for this purpose, enabling models to g…
Reinforcement LearningText GenerationDrug DiscoveryTemporal Alignment Guidance: On-Manifold Sampling in Diffusion Models
Diffusion models have achieved remarkable success as generative models. However, even a well-trained model can accumulate errors throughout the generation process. These errors become particularly problematic when arbitr…