Improving Pyramid Vector Quantizer with power projection
نویسنده
چکیده
Pyramid Vector Quantizer (PVQ) is a promising technique especially for multimedia data compression, already used in Opus audio codec and considered for AV1 video codec. It quantizes vectors from Euclidean unit sphere by first projecting them to L norm unit sphere, then quantizing and encoding there. This paper shows that the used standard radial projection is suboptimal and proposes to tune its deformations by using parameterized power projection: x → x/‖x‖ instead, where the optimized power p is applied coordinatewise, getting usually ≥ 0.5 dB improvement comparing to radial projection.
منابع مشابه
High-resolution quantization theory and the vector quantizer advantage
How much performance advantage can a fixed dimension vector quantizer gain over a scalar quantizer? We collect several results from high-resolution or asymptotic (in rate) quantization theory and use them to identify source and system characteristics that contribute to the vector quantizer advantage. One well-known advantage is due to improvement in the space-filling properties of polytopes as ...
متن کاملInterresolution Look-up Table for Improved Spatial Magnification of Image
The most commonly used image magnification techniques are interpolation based: nearest neighbor, bilinear, and bicubic. The drawbacks of these traditional methods are that images magnified by the simple nearest neighbor method often appear “blocky,” while images magnified by linear and cubic interpolations usually appear “blurry.” In this work, a new technique, which improves the performance of...
متن کاملA hybrid vector quantizer for enhanced image pyramid coding with application to volumetric image compression in confocal microscopy
Three-dimensional image compression methods outperform their two-dimensional counterparts in the sense of higher rate-distortion performance for compressing volumetric image data. The state-of-the-art transform-based 3D compressors, such as 3D-SPIHT and 3D-DCT, are characterized for their rate control ability, where the qualities of the image, although are adjustable with respect to rates, are ...
متن کامل3D Microscopic image coding by finite-state vector quantization in an enhanced image pyramid
A novel approach, based on a 3D difference pyramid structure with vector quantization error feedback, is proposed for microscopic volume image data compression. We have improved the coding performance relative to previous work. A finite-state vector quantizer (FSVQ) is introduced to exploit the correlation between neighbouring vectors to improve the coding efficiency. The effects of FSVQ in con...
متن کاملQuantizing Density Estimators
We suggest a nonparametric framework for unsupervised learning of projection models in terms of density estimation on quantized sample spaces. The objective is not to optimally reconstruct the data but instead the quantizer is chosen to optimally reconstruct the density of the data. For the resulting quantizing density estimator (QDE) we present a general method for parameter estimation and mod...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید
ثبت ناماگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید
ورودعنوان ژورنال:
- CoRR
دوره abs/1705.05285 شماره
صفحات -
تاریخ انتشار 2017