Towards Image Understanding from Deep Compression without Decoding

ثبت نشده
چکیده

Motivated by recent work on deep neural network (DNN)-based image compression methods showing potential improvements in image quality, savings in storage, and bandwidth reduction, we propose to perform image understanding tasks such as classification and segmentation directly on the compression representations produced by these compression methods. Since the encoders and decoders in DNN-based compression methods are neural networks with feature-maps as internal representations of the images, we directly integrate these with architectures for image understanding. This bypasses decoding of the compressed representation into RGB space and reduces computational cost. Our study shows that performance comparable to networks that operate on compressed RGB images can be achieved. Furthermore, we show that synergies are obtained by jointly training compression networks with classification networks on the compressed representations, improving image quality, classification accuracy, and segmentation performance. We find that inference from compressed representations is particularly advantageous compared to inference from compressed RGB images at high compression rates.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Towards Image Understanding from Deep Compression without Decoding

Motivated by recent work on deep neural network (DNN)-based image compression methods showing potential improvements in image quality, savings in storage, and bandwidth reduction, we propose to perform image understanding tasks such as classification and segmentation directly on the compressed representations produced by these compression methods. Since the encoders and decoders in DNN-based co...

متن کامل

ارائه روشی برای پیش‌پردازش تصویر جهت بهبود عملکرد JPEG

A lot of researchs have been performed in image compression and different methods have been proposed. Each of the existing methods presents different compression rates on various images. By identifing the effective parameters in a compression algorithm and strengthen them in the preprocessing stage, the compression rate of the algorithm can be improved. JPEG is one of the successful compression...

متن کامل

Bandwidth Efficient and Error Resilient Image Coding for Rayleigh Fading Channels

The transmission of still images over wireless channels is an exercise in tradeoos. For bandwidth eeciency, a high source compression rate is required. However, the transmission must be robust towards channel errors since wireless channels are subject to fading. These are connicting requirements. High source compression rates result in images that are very vulnerable to any single bit error. Lo...

متن کامل

DeepSIC: Deep Semantic Image Compression

Incorporating semantic information into the codecs during image compression can significantly reduce the repetitive computation of fundamental semantic analysis (such as object recognition) in client-side applications. The same practice also enable the compressed code to carry the image semantic information during storage and transmission. In this paper, we propose a concept called Deep Semanti...

متن کامل

A New Decoding Algorithm for Fractal Image Compression

A new iterative decoding method is proposed for fractal image compression. Convergence properties are provided. Experimental results show the superiority of the new method over the conventional decoding procedure.

متن کامل

ذخیره 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 نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2017