Data noising is an effective technique for regularizing neural network models. While noising is widely adopted in application domains such as vision and speech, commonly used noising primitives have not been developed for discrete sequencelevel settings such as language modeling. In this paper, we derive a connection between input noising in neural network language models and smoothing in ngram...

Translation invariant (TI) single wavelet de-noising was developed by Coifman and Donoho and they show that TI is better than non-TI single wavelet de-noising. On the other hand, Strela et al. have found that non-TI multiwavelet de-noising gives better results than non-TI single wavelets. In this paper we extend Coifman and Donoho's TI single wavelet de-noising scheme to multiwavelets. Experime...

De-noising is a substantial issue in hydrologic time series analysis, but it is a difficult task due to the defect of methods. In this paper an energy-based wavelet de-noising method was proposed. It is to remove noise by comparing energy distribution of series with the background energy distribution, which is established from Monte-Carlo test. Differing from wavelet threshold de-noising (WTD) ...

This paper briefly describes the basic principle of wavelet packet analysis, and on this basis introduces the general principle of wavelet packet transformation for signal den-noising. The dynamic EEG data under +Gz acceleration is made a de-noising treatment by using wavelet packet transformation, and the de-noising effects with different thresholds are made a comparison. The study verifies th...

Because the classic intersecting cortical model (ICM) and the traditional image de-noising algorithm exist the deficiencies-the image collection, transmission and conversion are often subjected to impulse noise interference, thus affecting the quality of the image, therefore we improved the framework structure and related parameters of the ICM and proposed the adaptive image de-noising algorith...

Signal de-noising is one of the classical problems in the field of signal processing. As a new signal processing tools, wavelet analysis, which has excellent noise performance, has caused growing concern and attention. The wavelet threshold de-noising has been researched systematacially, and the wavelet de-noising method is used on the GPS signal, which has achieved very good results.

Patch-based de-noising algorithms and patch manifold smoothing have emerged as efficient de-noising methods. This paper provides a new insight on these methods, such as the Non Local Means [1] or the image graph de-noising [8], by showing its use for filtering a selected pattern. K ̄ eywords: NL-Means, diffusion processes, diffusion geometry, graph filtering, patch manifold.

Image Preprocessing is a vital step in the field of image processing for biometric pattern recognition. This paper studies and reviews various classical and modern fingerprint image de-noising models. The various model used for de-noising ranges widely from transform matrix using frequency, histogram model denoising, de-noising by introducing Gabor filter and its types to enhance fingerprint im...

Over the past decade wavelet transforms have received a lot of attention from researchers in many diierent areas. Both discrete and continuous wavelet transforms have shown great promises in such diverse elds as image compression, image De-noising, signal processing, computer graphics, and pattern recognition, to name a few. Most of the work has been done on scalar wavelets, i.e., wavelets gene...