Context-Aware Selective Label Smoothing for Calibrating Sequence Recognition Model
This addresses the reliability issue in sequence recognition models for applications such as text and speech processing, though it is incremental as it builds on existing calibration methods.
The paper tackles the over-confidence problem in deep neural networks for sequence recognition by proposing a Context-Aware Selective Label Smoothing method, which leverages contextual dependencies to adaptively calibrate models, achieving state-of-the-art performance on tasks like scene text and speech recognition.
Despite the success of deep neural network (DNN) on sequential data (i.e., scene text and speech) recognition, it suffers from the over-confidence problem mainly due to overfitting in training with the cross-entropy loss, which may make the decision-making less reliable. Confidence calibration has been recently proposed as one effective solution to this problem. Nevertheless, the majority of existing confidence calibration methods aims at non-sequential data, which is limited if directly applied to sequential data since the intrinsic contextual dependency in sequences or the class-specific statistical prior is seldom exploited. To the end, we propose a Context-Aware Selective Label Smoothing (CASLS) method for calibrating sequential data. The proposed CASLS fully leverages the contextual dependency in sequences to construct confusion matrices of contextual prediction statistics over different classes. Class-specific error rates are then used to adjust the weights of smoothing strength in order to achieve adaptive calibration. Experimental results on sequence recognition tasks, including scene text recognition and speech recognition, demonstrate that our method can achieve the state-of-the-art performance.