This paper proposes a method that enhances the compression performance of the current model under development for the upcoming MPEG standard on Feature Compression for Machines (FCM) [1]. By truncating low-activation feature channels and signaling these truncations in the bitstream, the method reduces bitrate while preserving task accuracy. Experimental results show an average BD-rate reduction of 10.59% across datasets and tasks, demonstrating gains in bitrate efficiency.

Achieving successful variable bitrate compression with computationally simple algorithms from a single end-to-end learned image or video compression model remains a challenge. Many approaches have been proposed, including conditional auto-encoders, channel-adaptive gains for the latent tensor or uniformly quantizing all elements of the latent tensor. This paper follows the traditional approach to vary a single quantization step size to perform uniform quantization of all latent tensor elements. However, three modifications are proposed to improve the variable rate compression performance. First, multi objective optimization is used for (post) training. Second, a quantization-reconstruction offset is introduced into the quantization operation. Third, variable rate quantization is used also for the hyper latent. All these modifications can be made on a pre-trained single-rate compression model by performing post training. The algorithms are implemented into three well-known image compression models and the achieved variable rate compression results indicate negligible or minimal compression performance loss compared to training multiple models. (Codes will be shared at https://github.com/InterDigitalInc/CompressAI)

End-to-end image and video compression using auto-encoders (AE) offers new appealing perspectives in terms of rate-distortion gains and applications. While most complex models are on par with the latest compression standard like VVC/H.266 on objective metrics, practical implementation and complexity remain strong issues for real-world applications. We propose a practical implementation suitable for realistic applications. We demonstrate that some gains can be achieved on top low-complexity AE, even when using simpler implementation. The proposed implementation also allows a direct integration of such approaches on a variety of platforms and code is made available as a pure C++ standalone codec [1]:

As an increasing amount of image and video content will be analyzed by machines, there is demand for a new codec paradigm that is capable of compressing visual input primarily for the purpose of computer vision inference, while secondarily supporting input reconstruction. In this work, we propose a learned compression architecture that can be used to build such a codec. We introduce a novel variational formulation that explicitly takes feature data relevant to the desired inference task as input at the encoder side. As such, our learned scalable image codec encodes and transmits two disentangled latent representations for object detection and input reconstruction. We note that compared to relevant benchmarks, our proposed scheme yields a more compact latent representation that is specialized for the inference task. Our experiments show that our proposed system achieves a bit rate savings of 40.6% on the primary object detection task compared to the current state-of-the-art, albeit with some degradation in performance for the secondary input reconstruction task.

An increasing share of image and video content is analyzed by machines rather than viewed by humans, and therefore it becomes relevant to optimize codecs for such applications where the analysis is performed remotely. Unfortunately, conventional coding tools are challenging to specialize for machine tasks as they were originally designed for human perception. However, neural network based codecs can be jointly trained end-to-end with any convolutional neural network (CNN)-based task model. In this paper, we propose to study an end-to-end framework enabling efficient image compression for remote machine task analysis, using a chain composed of a compression module and a task algorithm that can be optimized end-to-end. We show that it is possible to significantly improve the task accuracy when fine-tuning jointly the codec and the task networks, especially at low bitrates. Depending on training or deployment constraints, selective fine-tuning can be applied only on the encoder, decoder or task network and still achieve rate-accuracy improvements over an off-the-shelf codec and task network. Our results also demonstrate the flexibility of end-to-end pipelines for practical applications.

Deep neural networks (DNNs), despite their performance on a wide variety of tasks, are still out of reach for many applications as they require significant computational resources. In this paper, we present a low-rank based end-to-end deep neural network compression framework with the goal of enabling DNNs performance to computationally constrained devices. The proposed framework includes techniques for low-rank based structural approximation, quantization and lossless arithmetic coding. Many of these techniques have been accepted in the MPEG working draft on compressed Neural Network Representations. We demonstrate the efficacy of the proposed framework via extensive experiments on a variety of DNNs for various tasks considered in this standardization activity. These techniques provide impressive performance on DNNs used in ImageNet Large-Scale Visual Recognition Challenge by compressing VGG16 by 61x, ResNet50 by almost 15x, and MobileNetV2 by almost 7x.

This paper provides a technical overview of a deep-learning-based encoder method aiming at optimizing next generation hybrid video encoders for driving the block partitioning in intra slices. An encoding approach based on Convolutional Neural Networks is explored to partly substitute classical heuristics-based encoder speed-ups by a systematic and automatic process. The solution allows controlling the trade-off between complexity and coding gains, in intra slices, with one single parameter. This algorithm was proposed at the Call for Proposals of the Joint Video Exploration Team (JVET) on video compression with capability beyond HEVC. In All Intra configuration, for a given allowed topology of splits, a speed-up of ×2 is obtained without BD-rate loss, or a speed-up above ×4 with a loss below 1% in BD-rate.

This paper presents a technical overview of Wide Angular Intra Prediction (WAIP) that was adopted into the test model of Versatile Video Coding (VVC) standard. Due to the adoption of flexible block partitioning using binary and ternary splits, a Coding Unit (CU) can have either a square or a rectangular block shape. However, the conventional angular intra prediction directions, ranging from 45 degrees to -135 degrees in clockwise direction, were designed for square CUs. To better optimize the intra prediction for rectangular blocks, WAIP modes were proposed to enable intra prediction directions beyond the range of conventional intra prediction directions. For different aspect ratios of rectangular block shapes, different number of conventional angular intra prediction modes were replaced by WAIP modes. The replaced intra prediction modes are signaled using the original signaling method. Simulation results reportedly show that, with almost no impact on the run-time, on average 0.31% BD-rate reduction is achieved for intra coding using VVC test model (VTM).

Summary form only given. This paper presents two sets of modifications to band offset type of the Sample Adaptive Offset technique in HEVC. First, some constraints on the SAO semantics are added to solve sub-optimal syntax issue and to exploit the actual range information of reconstructed samples. Next, the classification process is adapted to the particular samples statistics.

This paper presents an adaptive clipping technique with optimized syntax in the video coding Joint Exploratory Model (JEM), which exploits the signal characteristics of the video sequence. The component-wise clipping bounds are coded for each slice. Two encoding methods leveraging the efficiency of the proposed technique are then described. The first one consists in modeling the errors induced by the clipping process in the Rate Distortion Optimization. The second one aims at reducing the cost of transform coefficients by smoothing the residuals. Finally, experimental results are provided and several variants are discussed.