Today, according to the Cisco Annual Internet Report (2018-2023), the fastest-growing category of Internet traffic is machine-to-machine communication. In particular, machine-to-machine communication of images and videos represents a new challenge and opens up new perspectives in the context of data compression. One possible solution approach consists of adapting current human-targeted image and video coding standards to the use case of machine consumption. Another approach consists of developing completely new compression paradigms and architectures for machine-to-machine communications. In this paper, we focus on image compression and present an inference-time content-adaptive fine-tuning scheme that optimizes the latent representation of an end-to-end learned image codec, aimed at improving the compression efficiency for machine-consumption. The conducted experiments targeting instance segmentation task network show that our online finetuning brings an average bitrate saving (BD-rate) of -3.66% with respect to our pretrained image codec. In particular, at low bitrate points, our proposed method results in a significant bitrate saving of -9.85%. Overall, our pretrained-and-then-finetuned system achieves - 30.54% BD-rate over the state-of-the-art image/video codec Versatile Video Coding (VVC) on instance segmentation.

Learned Image Coding for Machines: A Content-Adaptive Approach

In this paper we present an end-to-end meta-learned system for image compression. Traditional machine learning based approaches to image compression train one or more neural network for generalization performance. However, at inference time, the encoder or the latent tensor output by the encoder can be optimized for each test image. This optimization can be regarded as a form of adaptation or benevolent overfitting to the input content. In order to reduce the gap between training and inference conditions, we propose a new training paradigm for learned image compression, which is based on meta-learning. In a first phase, the neural networks are trained normally. In a second phase, the Model-Agnostic Meta-learning approach is adapted to the specific case of image compression, where the inner-loop performs latent tensor overfitting, and the outer loop updates both encoder and decoder neural networks based on the overfitting performance. Furthermore, after meta-learning, we propose to overfit and cluster the bias terms of the decoder on training image patches, so that at inference time the optimal content-specific bias terms can be selected at encoder-side. Finally, we propose a new probability model for lossless compression, which combines concepts from both multi-scale and super-resolution probability model approaches. We show the benefits of all our proposed ideas via carefully designed experiments.

L 2 C – Learning to Learn to Compress

L$^2$C -- Learning to Learn to Compress

Screen content data, such as computer-generated photographs, desktop sharing, remote education, video game streaming and screenshot, is one of the most popular visual information carriers in Internet of Video Things. Although lossless compression can guarantee high quality of service for these screen content based industrial applications, it also causes considerable storage space and transmission bandwidth issues. To alleviate these challenges, in this article, we present a visually quasi-lossless coding approach to control the compression distortion below visibility threshold in the human visual system. Specifically, to better quantify the visual redundancy for screen content data, a new visibility threshold method is designed by incorporating blur sensitivity and oblique correction effects. Then, an end-to-end mapping between the visibility threshold and quality control factor is learned and represented as a deep convolutional neural network. The experimental results demonstrate that the proposed method saves the average encoding bits up to 23.15% compared with the latest scheme under the same perceptual quality.

Perceptually Quasi-Lossless Compression of Screen Content Data Via Visibility Modeling and Deep Forecasting

Lossy image compression is a popular, simple and effective solution to reduce the amount of data representing digital pictures. In most lossy compression methods, the reduced volume of data in bits is achieved at the expense of introducing visual artifacts in the picture. The perceptual quality impact of such artifacts can be assessed with expensive and time-consuming subjective image quality experiments or through objective image quality metrics. However, the faster and less resource demanding objective quality metrics are not always able to reliably predict the quality as perceived by human observers. In this paper, the performance of 14 objective image quality metrics is benchmarked against a dataset of compressed images labeled with their subjective quality scores. Moreover, the performance of the above objective quality metrics in predicting the subjective quality of images distorted by both conventional and learning-based lossy compression artifacts is assessed and conclusions are drawn.

/pdf/performance-evaluation-of-objective-image-quality-metrics-on-3c49bt7ahp.pdf

Performance Evaluation of Objective Image Quality Metrics on Conventional and Learning-Based Compression Artifacts

Lossless image compression is an important technique for image storage and transmission when information loss is not allowed. With the fast development of deep learning techniques, deep neural networks have been used in this field to achieve a higher compression rate. Methods based on pixel-wise autoregressive statistical models have shown good performance. However, the sequential processing way prevents these methods to be used in practice. Recently, multi-scale autoregressive models have been proposed to address this limitation. Multi-scale approaches can use parallel computing systems efficiently and build practical systems. Nevertheless, these approaches sacrifice compression performance in exchange for speed. In this paper, we propose a multi-scale progressive statistical model that takes advantage of the pixel-wise approach and the multi-scale approach. We developed a flexible mechanism where the processing order of the pixels can be adjusted easily. Our proposed method outperforms the state-of-the-art lossless image compression methods on two large benchmark datasets by a significant margin without degrading the inference speed dramatically.

/pdf/lossless-image-compression-using-a-multi-scale-progressive-3ilzu0v2t9.pdf

Lossless Image Compression Using a Multi-Scale Progressive Statistical Model

One of the core components of conventional (i.e., non-learned) video codecs consists of predicting a frame from a previously-decoded frame, by leveraging temporal correlations. In this paper, we propose an end-to-end learned system for compressing video frames. Instead of relying on pixel-space motion (as with optical flow), our system learns deep embeddings of frames and encodes their difference in latent space. At decoder-side, an attention mechanism is designed to attend to the latent space of frames to decide how different parts of the previous and current frame are combined to form the final predicted current frame. Spatially-varying channel allocation is achieved by using importance masks acting on the feature-channels. The model is trained to reduce the bitrate by minimizing a loss on importance maps and a loss on the probability output by a context model for arithmetic coding. In our experiments, we show that the proposed system achieves high compression rates and high objective visual quality as measured by MS-SSIM and PSNR. Furthermore, we provide ablation studies where we highlight the contribution of different components.

/pdf/end-to-end-learning-for-video-frame-compression-with-self-1jc2249htd.pdf

Nannan Zou

Papers

L 2 C – Learning to Learn to Compress

Lossless Image Compression Using a Multi-Scale Progressive Statistical Model

L$^2$C -- Learning to Learn to Compress

End-to-End Learning for Video Frame Compression with Self-Attention

End-to-End Learning for Video Frame Compression with Self-Attention