Despite considerable progress on end-to-end optimized deep networks for image compression, video coding remains a challenging task. Recently proposed methods for learned video compression use optical flow and bilinear warping for motion compensation and show competitive rate-distortion performance relative to hand-engineered codecs like H.264 and HEVC. However, these learning-based methods rely on complex architectures and training schemes including the use of pre-trained optical flow networks, sequential training of sub-networks, adaptive rate control, and buffering intermediate reconstructions to disk during training. In this paper, we show that a generalized warping operator that better handles common failure cases, e.g. disocclusions and fast motion, can provide competitive compression results with a greatly simplified model and training procedure. Specifically, we propose scale-space flow, an intuitive generalization of optical flow that adds a scale parameter to allow the network to better model uncertainty. Our experiments show that a low-latency video compression model (no B-frames) using scale-space flow for motion compensation can outperform analogous state-of-the art learned video compression models while being trained using a much simpler procedure and without any pre-trained optical flow networks.

/pdf/scale-space-flow-for-end-to-end-optimized-video-compression-2lvsrtyw3y.pdf

Scale-Space Flow for End-to-End Optimized Video Compression

We propose a neural talking-head video synthesis model and demonstrate its application to video conferencing. Our model learns to synthesize a talking-head video using a source image containing the target person's appearance and a driving video that dictates the motion in the output. Our motion is encoded based on a novel keypoint representation, where the identity-specific and motion-related information is decomposed unsupervisedly. Extensive experimental validation shows that our model outperforms competing methods on benchmark datasets. Moreover, our compact keypoint representation enables a video conferencing system that achieves the same visual quality as the commercial H.264 standard while only using one-tenth of the bandwidth. Besides, we show our keypoint representation allows the user to rotate the head during synthesis, which is useful for simulating face-to-face video conferencing experiences.

One-Shot Free-View Neural Talking-Head Synthesis for Video Conferencing

In this paper we present a a deep generative model for lossy video compression. We employ a model that consists of a 3D autoencoder with a discrete latent space and an autoregressive prior used for entropy coding. Both autoencoder and prior are trained jointly to minimize a rate-distortion loss, which is closely related to the ELBO used in variational autoencoders. Despite its simplicity, we find that our method outperforms the state-of-the-art learned video compression networks based on motion compensation or interpolation. We systematically evaluate various design choices, such as the use of frame-based or spatio-temporal autoencoders, and the type of autoregressive prior. In addition, we present three extensions of the basic method that demonstrate the benefits over classical approaches to compression. First, we introduce semantic compression, where the model is trained to allocate more bits to objects of interest. Second, we study adaptive compression, where the model is adapted to a domain with limited variability, e.g., videos taken from an autonomous car, to achieve superior compression on that domain. Finally, we introduce multimodal compression, where we demonstrate the effectiveness of our model in joint compression of multiple modalities captured by non-standard imaging sensors, such as quad cameras. We believe that this opens up novel video compression applications, which have not been feasible with classical codecs.

Video Compression With Rate-Distortion Autoencoders

We review a class of methods that can be collected under the name nonlinear transform coding (NTC), which over the past few years have become competitive with the best linear transform codecs for images, and have superseded them in terms of rate–distortion performance under established perceptual quality metrics such as MS-SSIM. We assess the empirical rate–distortion performance of NTC with the help of simple example sources, for which the optimal performance of a vector quantizer is easier to estimate than with natural data sources. To this end, we introduce a novel variant of entropy-constrained vector quantization. We provide an analysis of various forms of stochastic optimization techniques for NTC models; review architectures of transforms based on artificial neural networks, as well as learned entropy models; and provide a direct comparison of a number of methods to parameterize the rate–distortion trade-off of nonlinear transforms, introducing a simplified one.

/pdf/nonlinear-transform-coding-402wbjdyjh.pdf

Nonlinear Transform Coding

Traditional video compression approaches build upon the hybrid coding framework with motion-compensated prediction and residual transform coding. In this paper, we propose the first end-to-end deep video compression framework to take advantage of both the classical compression architecture and the powerful non-linear representation ability of neural networks. Our framework employs pixel-wise motion information, which is learned from an optical flow network and further compressed by an auto-encoder network to save bits. The other compression components are also implemented by the well-designed networks for high efficiency. All the modules are jointly optimized by using the rate-distortion trade-off and can collaborate with each other. More importantly, the proposed deep video compression framework is very flexible and can be easily extended by using lightweight or advanced networks for higher speed or better efficiency. We also propose to introduce the adaptive quantization layer to reduce the number of parameters for variable bitrate coding. Comprehensive experimental results demonstrate the effectiveness of the proposed framework on the benchmark datasets.

An End-to-End Learning Framework for Video Compression

We present a new algorithm for video coding, learned end-to-end for the low-latency mode. In this setting, our approach outperforms all existing video codecs across nearly the entire bitrate range. To our knowledge, this is the first ML-based method to do so. We evaluate our approach on standard video compression test sets of varying resolutions, and benchmark against all mainstream commercial codecs in the low-latency mode. On standard-definition videos, HEVC/H.265, AVC/H.264 and VP9 typically produce codes up to 60% larger than our algorithm. On high-definition 1080p videos, H.265 and VP9 typically produce codes up to 20% larger, and H.264 up to 35% larger. Furthermore, our approach does not suffer from blocking artifacts and pixelation, and thus produces videos that are more visually pleasing. We propose two main contributions. The first is a novel architecture for video compression, which (1) generalizes motion estimation to perform any learned compensation beyond simple translations, (2) rather than strictly relying on previously transmitted reference frames, maintains a state of arbitrary information learned by the model, and (3) enables jointly compressing all transmitted signals (such as optical flow and residual). Secondly, we present a framework for ML-based spatial rate control --- a mechanism for assigning variable bitrates across space for each frame. This is a critical component for video coding, which to our knowledge had not been developed within a machine learning setting.

/pdf/learned-video-compression-5e2mmrgtey.pdf

Learned Video Compression

We present a new algorithm for video coding, learned end-to-end for the low-latency mode. In this setting, our approach outperforms all existing video codecs across nearly the entire bitrate range. To our knowledge, this is the first ML-based method to do so. 
We evaluate our approach on standard video compression test sets of varying resolutions, and benchmark against all mainstream commercial codecs, in the low-latency mode. On standard-definition videos, relative to our algorithm, HEVC/H.265, AVC/H.264 and VP9 typically produce codes up to 60% larger. On high-definition 1080p videos, H.265 and VP9 typically produce codes up to 20% larger, and H.264 up to 35% larger. Furthermore, our approach does not suffer from blocking artifacts and pixelation, and thus produces videos that are more visually pleasing. 
We propose two main contributions. The first is a novel architecture for video compression, which (1) generalizes motion estimation to perform any learned compensation beyond simple translations, (2) rather than strictly relying on previously transmitted reference frames, maintains a state of arbitrary information learned by the model, and (3) enables jointly compressing all transmitted signals (such as optical flow and residual). 
Secondly, we present a framework for ML-based spatial rate control: namely, a mechanism for assigning variable bitrates across space for each frame. This is a critical component for video coding, which to our knowledge had not been developed within a machine learning setting.

The compression system trains a machine-learned encoder and decoder through an autoencoder architecture. The encoder can be deployed by a sender system to encode content for transmission to a receiver system, and the decoder can be deployed by the receiver system to decode the encoded content and reconstruct the original content. The encoder is coupled to receive content and output a tensor as a compact representation of the content. The content may be, for example, images, videos, or text. The decoder is coupled to receive a tensor representing content and output a reconstructed version of the content. The compression system trains the autoencoder with a discriminator to reduce compression artifacts in the reconstructed content. The discriminator is coupled to receive one or more input content, and output a discrimination prediction that discriminates whether the input content is the original or reconstructed version of the content.

Using generative adversarial networks in compression

An enhanced encoder system generates residual bitstreams representing additional image information that can be used by an image enhancement system to improve a low quality image. The enhanced encoder system upsamples a low quality image and compares the upsampled image to a true high quality image to determine image inaccuracies that arise due to the upsampling process. The enhanced encoder system encodes the information describing the image inaccuracies using a trained encoder model as the residual bitstream. The image enhancement system upsamples the same low quality image to obtain a prediction of a high quality image that can include image inaccuracies. Given the residual bitstream, the image enhancement system decodes the residual bitstream using a trained decoder model and uses the additional image information to improve the predicted high quality image. The image enhancement system can provide an improved, high quality image for display.

Autoencoding image residuals for improving upsampled images

A compression system trains a machine-learned encoder and decoder. The encoder can be deployed by a sender system to encode content for transmission to a receiver system, and the decoder can be deployed by the receiver system to decode the encoded content and reconstruct the original content. The encoder receives content and generates a tensor as a compact representation of the content. The content may be, for example, images, videos, or text. The decoder receives a tensor and generates a reconstructed version of the content. In one embodiment, the compression system trains one or more encoding components such that the encoder can adaptively encode different degrees of information for regions in the content that are associated with characteristic objects, such as human faces, texts, or buildings.

Sanjay Nair

Papers

Learned Video Compression

Learned Video Compression

Using generative adversarial networks in compression

Autoencoding image residuals for improving upsampled images

Adaptive compression based on content