George Toderici

Fabian Mentzer, George Toderici, David Minnen et al.

We show how transformers can be used to vastly simplify neural video compression. Previous methods have been relying on an increasing number of architectural biases and priors, including motion prediction and warping operations, resulting in complex models. Instead, we independently map input frames to representations and use a transformer to model their dependencies, letting it predict the distribution of future representations given the past. The resulting video compression transformer outperforms previous methods on standard video compression data sets. Experiments on synthetic data show that our model learns to handle complex motion patterns such as panning, blurring and fading purely from data. Our approach is easy to implement, and we release code to facilitate future research.

Eirikur Agustsson, David Minnen, George Toderici et al.

By optimizing the rate-distortion-realism trade-off, generative compression approaches produce detailed, realistic images, even at low bit rates, instead of the blurry reconstructions produced by rate-distortion optimized models. However, previous methods do not explicitly control how much detail is synthesized, which results in a common criticism of these methods: users might be worried that a misleading reconstruction far from the input image is generated. In this work, we alleviate these concerns by training a decoder that can bridge the two regimes and navigate the distortion-realism trade-off. From a single compressed representation, the receiver can decide to either reconstruct a low mean squared error reconstruction that is close to the input, a realistic reconstruction with high perceptual quality, or anything in between. With our method, we set a new state-of-the-art in distortion-realism, pushing the frontier of achievable distortion-realism pairs, i.e., our method achieves better distortions at high realism and better realism at low distortion than ever before.

Robert Geirhos, Priyank Jaini, Austin Stone et al.

Training a neural network is a monolithic endeavor, akin to carving knowledge into stone: once the process is completed, editing the knowledge in a network is hard, since all information is distributed across the network's weights. We here explore a simple, compelling alternative by marrying the representational power of deep neural networks with the flexibility of a database. Decomposing the task of image classification into image similarity (from a pre-trained embedding) and search (via fast nearest neighbor retrieval from a knowledge database), we build on well-established components to construct a simple and flexible visual memory that has the following key capabilities: (1.) The ability to flexibly add data across scales: from individual samples all the way to entire classes and billion-scale data; (2.) The ability to remove data through unlearning and memory pruning; (3.) An interpretable decision-mechanism on which we can intervene to control its behavior. Taken together, these capabilities comprehensively demonstrate the benefits of an explicit visual memory. We hope that it might contribute to a conversation on how knowledge should be represented in deep vision models -- beyond carving it in "stone" weights.

Sami Abu-El-Haija, Nisarg Kothari, Joonseok Lee et al.

Many recent advancements in Computer Vision are attributed to large datasets. Open-source software packages for Machine Learning and inexpensive commodity hardware have reduced the barrier of entry for exploring novel approaches at scale. It is possible to train models over millions of examples within a few days. Although large-scale datasets exist for image understanding, such as ImageNet, there are no comparable size video classification datasets. In this paper, we introduce YouTube-8M, the largest multi-label video classification dataset, composed of ~8 million videos (500K hours of video), annotated with a vocabulary of 4800 visual entities. To get the videos and their labels, we used a YouTube video annotation system, which labels videos with their main topics. While the labels are machine-generated, they have high-precision and are derived from a variety of human-based signals including metadata and query click signals. We filtered the video labels (Knowledge Graph entities) using both automated and manual curation strategies, including asking human raters if the labels are visually recognizable. Then, we decoded each video at one-frame-per-second, and used a Deep CNN pre-trained on ImageNet to extract the hidden representation immediately prior to the classification layer. Finally, we compressed the frame features and make both the features and video-level labels available for download. We trained various (modest) classification models on the dataset, evaluated them using popular evaluation metrics, and report them as baselines. Despite the size of the dataset, some of our models train to convergence in less than a day on a single machine using TensorFlow. We plan to release code for training a TensorFlow model and for computing metrics.

Emiel Hoogeboom, Eirikur Agustsson, Fabian Mentzer et al.

Despite the tremendous success of diffusion generative models in text-to-image generation, replicating this success in the domain of image compression has proven difficult. In this paper, we demonstrate that diffusion can significantly improve perceptual quality at a given bit-rate, outperforming state-of-the-art approaches PO-ELIC and HiFiC as measured by FID score. This is achieved using a simple but theoretically motivated two-stage approach combining an autoencoder targeting MSE followed by a further score-based decoder. However, as we will show, implementation details matter and the optimal design decisions can differ greatly from typical text-to-image models.

Berivan Isik, Philip A. Chou, Sung Jin Hwang et al.

We consider the attributes of a point cloud as samples of a vector-valued volumetric function at discrete positions. To compress the attributes given the positions, we compress the parameters of the volumetric function. We model the volumetric function by tiling space into blocks, and representing the function over each block by shifts of a coordinate-based, or implicit, neural network. Inputs to the network include both spatial coordinates and a latent vector per block. We represent the latent vectors using coefficients of the region-adaptive hierarchical transform (RAHT) used in the MPEG geometry-based point cloud codec G-PCC. The coefficients, which are highly compressible, are rate-distortion optimized by back-propagation through a rate-distortion Lagrangian loss in an auto-decoder configuration. The result outperforms RAHT by 2--4 dB. This is the first work to compress volumetric functions represented by local coordinate-based neural networks. As such, we expect it to be applicable beyond point clouds, for example to compression of high-resolution neural radiance fields.

Fabian Mentzer, Eirikur Agustsson, Johannes Ballé et al.

We present the first neural video compression method based on generative adversarial networks (GANs). Our approach significantly outperforms previous neural and non-neural video compression methods in a user study, setting a new state-of-the-art in visual quality for neural methods. We show that the GAN loss is crucial to obtain this high visual quality. Two components make the GAN loss effective: we i) synthesize detail by conditioning the generator on a latent extracted from the warped previous reconstruction to then ii) propagate this detail with high-quality flow. We find that user studies are required to compare methods, i.e., none of our quantitative metrics were able to predict all studies. We present the network design choices in detail, and ablate them with user studies.

Saurabh Singh, Sami Abu-El-Haija, Nick Johnston et al.

Pre-trained convolutional neural networks (CNNs) are powerful off-the-shelf feature generators and have been shown to perform very well on a variety of tasks. Unfortunately, the generated features are high dimensional and expensive to store: potentially hundreds of thousands of floats per example when processing videos. Traditional entropy based lossless compression methods are of little help as they do not yield desired level of compression, while general purpose lossy compression methods based on energy compaction (e.g. PCA followed by quantization and entropy coding) are sub-optimal, as they are not tuned to task specific objective. We propose a learned method that jointly optimizes for compressibility along with the task objective for learning the features. The plug-in nature of our method makes it straight-forward to integrate with any target objective and trade-off against compressibility. We present results on multiple benchmarks and demonstrate that our method produces features that are an order of magnitude more compressible, while having a regularization effect that leads to a consistent improvement in accuracy.

Fabian Mentzer, George Toderici, Michael Tschannen et al.

We extensively study how to combine Generative Adversarial Networks and learned compression to obtain a state-of-the-art generative lossy compression system. In particular, we investigate normalization layers, generator and discriminator architectures, training strategies, as well as perceptual losses. In contrast to previous work, i) we obtain visually pleasing reconstructions that are perceptually similar to the input, ii) we operate in a broad range of bitrates, and iii) our approach can be applied to high-resolution images. We bridge the gap between rate-distortion-perception theory and practice by evaluating our approach both quantitatively with various perceptual metrics, and with a user study. The study shows that our method is preferred to previous approaches even if they use more than 2x the bitrate.

David Minnen, Johannes Ballé, George Toderici

Recent models for learned image compression are based on autoencoders, learning approximately invertible mappings from pixels to a quantized latent representation. These are combined with an entropy model, a prior on the latent representation that can be used with standard arithmetic coding algorithms to yield a compressed bitstream. Recently, hierarchical entropy models have been introduced as a way to exploit more structure in the latents than simple fully factorized priors, improving compression performance while maintaining end-to-end optimization. Inspired by the success of autoregressive priors in probabilistic generative models, we examine autoregressive, hierarchical, as well as combined priors as alternatives, weighing their costs and benefits in the context of image compression. While it is well known that autoregressive models come with a significant computational penalty, we find that in terms of compression performance, autoregressive and hierarchical priors are complementary and, together, exploit the probabilistic structure in the latents better than all previous learned models. The combined model yields state-of-the-art rate--distortion performance, providing a 15.8% average reduction in file size over the previous state-of-the-art method based on deep learning, which corresponds to a 59.8% size reduction over JPEG, more than 35% reduction compared to WebP and JPEG2000, and bitstreams 8.4% smaller than BPG, the current state-of-the-art image codec. To the best of our knowledge, our model is the first learning-based method to outperform BPG on both PSNR and MS-SSIM distortion metrics.

Troy Chinen, Johannes Ballé, Chunhui Gu et al.

We present a full reference, perceptual image metric based on VGG-16, an artificial neural network trained on object classification. We fit the metric to a new database based on 140k unique images annotated with ground truth by human raters who received minimal instruction. The resulting metric shows competitive performance on TID 2013, a database widely used to assess image quality assessments methods. More interestingly, it shows strong responses to objects potentially carrying semantic relevance such as faces and text, which we demonstrate using a visualization technique and ablation experiments. In effect, the metric appears to model a higher influence of semantic context on judgments, which we observe particularly in untrained raters. As the vast majority of users of image processing systems are unfamiliar with Image Quality Assessment (IQA) tasks, these findings may have significant impact on real-world applications of perceptual metrics.

David Minnen, George Toderici, Saurabh Singh et al.

The leading approach for image compression with artificial neural networks (ANNs) is to learn a nonlinear transform and a fixed entropy model that are optimized for rate-distortion performance. We show that this approach can be significantly improved by incorporating spatially local, image-dependent entropy models. The key insight is that existing ANN-based methods learn an entropy model that is shared between the encoder and decoder, but they do not transmit any side information that would allow the model to adapt to the structure of a specific image. We present a method for augmenting ANN-based image coders with image-dependent side information that leads to a 17.8% rate reduction over a state-of-the-art ANN-based baseline model on a standard evaluation set, and 70-98% reductions on images with low visual complexity that are poorly captured by a fixed, global entropy model.

David Minnen, George Toderici, Michele Covell et al.

Deep neural networks represent a powerful class of function approximators that can learn to compress and reconstruct images. Existing image compression algorithms based on neural networks learn quantized representations with a constant spatial bit rate across each image. While entropy coding introduces some spatial variation, traditional codecs have benefited significantly by explicitly adapting the bit rate based on local image complexity and visual saliency. This paper introduces an algorithm that combines deep neural networks with quality-sensitive bit rate adaptation using a tiled network. We demonstrate the importance of spatial context prediction and show improved quantitative (PSNR) and qualitative (subjective rater assessment) results compared to a non-adaptive baseline and a recently published image compression model based on fully-convolutional neural networks.

Chunhui Gu, Chen Sun, David A. Ross et al.

This paper introduces a video dataset of spatio-temporally localized Atomic Visual Actions (AVA). The AVA dataset densely annotates 80 atomic visual actions in 430 15-minute video clips, where actions are localized in space and time, resulting in 1.58M action labels with multiple labels per person occurring frequently. The key characteristics of our dataset are: (1) the definition of atomic visual actions, rather than composite actions; (2) precise spatio-temporal annotations with possibly multiple annotations for each person; (3) exhaustive annotation of these atomic actions over 15-minute video clips; (4) people temporally linked across consecutive segments; and (5) using movies to gather a varied set of action representations. This departs from existing datasets for spatio-temporal action recognition, which typically provide sparse annotations for composite actions in short video clips. We will release the dataset publicly. AVA, with its realistic scene and action complexity, exposes the intrinsic difficulty of action recognition. To benchmark this, we present a novel approach for action localization that builds upon the current state-of-the-art methods, and demonstrates better performance on JHMDB and UCF101-24 categories. While setting a new state of the art on existing datasets, the overall results on AVA are low at 15.6% mAP, underscoring the need for developing new approaches for video understanding.

Michele Covell, Nick Johnston, David Minnen et al.

We introduce a stop-code tolerant (SCT) approach to training recurrent convolutional neural networks for lossy image compression. Our methods introduce a multi-pass training method to combine the training goals of high-quality reconstructions in areas around stop-code masking as well as in highly-detailed areas. These methods lead to lower true bitrates for a given recursion count, both pre- and post-entropy coding, even using unstructured LZ77 code compression. The pre-LZ77 gains are achieved by trimming stop codes. The post-LZ77 gains are due to the highly unequal distributions of 0/1 codes from the SCT architectures. With these code compressions, the SCT architecture maintains or exceeds the image quality at all compression rates compared to JPEG and to RNN auto-encoders across the Kodak dataset. In addition, the SCT coding results in lower variance in image quality across the extent of the image, a characteristic that has been shown to be important in human ratings of image quality

Nick Johnston, Damien Vincent, David Minnen et al.

We propose a method for lossy image compression based on recurrent, convolutional neural networks that outperforms BPG (4:2:0 ), WebP, JPEG2000, and JPEG as measured by MS-SSIM. We introduce three improvements over previous research that lead to this state-of-the-art result. First, we show that training with a pixel-wise loss weighted by SSIM increases reconstruction quality according to several metrics. Second, we modify the recurrent architecture to improve spatial diffusion, which allows the network to more effectively capture and propagate image information through the network's hidden state. Finally, in addition to lossless entropy coding, we use a spatially adaptive bit allocation algorithm to more efficiently use the limited number of bits to encode visually complex image regions. We evaluate our method on the Kodak and Tecnick image sets and compare against standard codecs as well recently published methods based on deep neural networks.

George Toderici, Damien Vincent, Nick Johnston et al.

This paper presents a set of full-resolution lossy image compression methods based on neural networks. Each of the architectures we describe can provide variable compression rates during deployment without requiring retraining of the network: each network need only be trained once. All of our architectures consist of a recurrent neural network (RNN)-based encoder and decoder, a binarizer, and a neural network for entropy coding. We compare RNN types (LSTM, associative LSTM) and introduce a new hybrid of GRU and ResNet. We also study "one-shot" versus additive reconstruction architectures and introduce a new scaled-additive framework. We compare to previous work, showing improvements of 4.3%-8.8% AUC (area under the rate-distortion curve), depending on the perceptual metric used. As far as we know, this is the first neural network architecture that is able to outperform JPEG at image compression across most bitrates on the rate-distortion curve on the Kodak dataset images, with and without the aid of entropy coding.

George Toderici, Sean M. O'Malley, Sung Jin Hwang et al.

A large fraction of Internet traffic is now driven by requests from mobile devices with relatively small screens and often stringent bandwidth requirements. Due to these factors, it has become the norm for modern graphics-heavy websites to transmit low-resolution, low-bytecount image previews (thumbnails) as part of the initial page load process to improve apparent page responsiveness. Increasing thumbnail compression beyond the capabilities of existing codecs is therefore a current research focus, as any byte savings will significantly enhance the experience of mobile device users. Toward this end, we propose a general framework for variable-rate image compression and a novel architecture based on convolutional and deconvolutional LSTM recurrent networks. Our models address the main issues that have prevented autoencoder neural networks from competing with existing image compression algorithms: (1) our networks only need to be trained once (not per-image), regardless of input image dimensions and the desired compression rate; (2) our networks are progressive, meaning that the more bits are sent, the more accurate the image reconstruction; and (3) the proposed architecture is at least as efficient as a standard purpose-trained autoencoder for a given number of bits. On a large-scale benchmark of 32$\times$32 thumbnails, our LSTM-based approaches provide better visual quality than (headerless) JPEG, JPEG2000 and WebP, with a storage size that is reduced by 10% or more.

Greg Mori, Caroline Pantofaru, Nisarg Kothari et al.

We present a method for learning an embedding that places images of humans in similar poses nearby. This embedding can be used as a direct method of comparing images based on human pose, avoiding potential challenges of estimating body joint positions. Pose embedding learning is formulated under a triplet-based distance criterion. A deep architecture is used to allow learning of a representation capable of making distinctions between different poses. Experiments on human pose matching and retrieval from video data demonstrate the potential of the method.

Balakrishnan Varadarajan, George Toderici, Sudheendra Vijayanarasimhan et al.

Video classification has advanced tremendously over the recent years. A large part of the improvements in video classification had to do with the work done by the image classification community and the use of deep convolutional networks (CNNs) which produce competitive results with hand- crafted motion features. These networks were adapted to use video frames in various ways and have yielded state of the art classification results. We present two methods that build on this work, and scale it up to work with millions of videos and hundreds of thousands of classes while maintaining a low computational cost. In the context of large scale video processing, training CNNs on video frames is extremely time consuming, due to the large number of frames involved. We propose to avoid this problem by training CNNs on either YouTube thumbnails or Flickr images, and then using these networks' outputs as features for other higher level classifiers. We discuss the challenges of achieving this and propose two models for frame-level and video-level classification. The first is a highly efficient mixture of experts while the latter is based on long short term memory neural networks. We present results on the Sports-1M video dataset (1 million videos, 487 classes) and on a new dataset which has 12 million videos and 150,000 labels.

Joe Yue-Hei Ng, Matthew Hausknecht, Sudheendra Vijayanarasimhan et al.

Convolutional neural networks (CNNs) have been extensively applied for image recognition problems giving state-of-the-art results on recognition, detection, segmentation and retrieval. In this work we propose and evaluate several deep neural network architectures to combine image information across a video over longer time periods than previously attempted. We propose two methods capable of handling full length videos. The first method explores various convolutional temporal feature pooling architectures, examining the various design choices which need to be made when adapting a CNN for this task. The second proposed method explicitly models the video as an ordered sequence of frames. For this purpose we employ a recurrent neural network that uses Long Short-Term Memory (LSTM) cells which are connected to the output of the underlying CNN. Our best networks exhibit significant performance improvements over previously published results on the Sports 1 million dataset (73.1% vs. 60.9%) and the UCF-101 datasets with (88.6% vs. 88.0%) and without additional optical flow information (82.6% vs. 72.8%).