Thomas Maugey

Caroline Mazini Rodrigues, Nicolas Keriven, Thomas Maugey

Deep learning models for image compression often face practical limitations in hardware-constrained applications. Although these models achieve high-quality reconstructions, they are typically complex, heavyweight, and require substantial training data and computational resources. We propose a methodology to partially compress these networks by reducing the size of their encoders. Our approach uses a simplified knowledge distillation strategy to approximate the latent space of the original models with less data and shorter training, yielding lightweight encoders from heavyweight ones. We evaluate the resulting lightweight encoders across two different architectures on the image compression task. Experiments show that our method preserves reconstruction quality and statistical fidelity better than training lightweight encoders with the original loss, making it practical for resource-limited environments.

Esteban Pesnel, Julien Le Tanou, Michael Ropert et al.

The rapid growth in video consumption has introduced significant challenges to modern streaming architectures. Over-the-Top (OTT) video delivery now predominantly relies on Adaptive Bitrate (ABR) streaming, which dynamically adjusts bitrate and resolution based on client-side constraints such as display capabilities and network bandwidth. This pipeline typically involves downsampling the original high-resolution content, encoding and transmitting it, followed by decoding and upsampling on the client side. Traditionally, these processing stages have been optimized in isolation, leading to suboptimal end-to-end rate-distortion (R-D) performance. The advent of deep learning has spurred interest in jointly optimizing the ABR pipeline using learned resampling methods. However, training such systems end-to-end remains challenging due to the non-differentiable nature of standard video codecs, which obstructs gradient-based optimization. Recent works have addressed this issue using differentiable proxy models, based either on deep neural networks or hybrid coding schemes with differentiable components such as soft quantization, to approximate the codec behavior. While differentiable proxy codecs have enabled progress in compression-aware learning, they remain approximations that may not fully capture the behavior of standard, non-differentiable codecs. To our knowledge, there is no prior evidence demonstrating the inefficiencies of using standard codecs during training. In this work, we introduce a novel framework that enables end-to-end training with real, non-differentiable codecs by leveraging data-driven surrogate gradients derived from actual compression errors. It facilitates the alignment between training objectives and deployment performance. Experimental results show a 5.19\% improvement in BD-BR (PSNR) compared to codec-agnostic training approaches, consistently across the entire rate-distortion convex hull spanning multiple downsampling ratios.

Tom Bordin, Thomas Maugey

This study addresses the challenge of, without training or fine-tuning, controlling the global color aspect of images generated with a diffusion model. We rewrite the guidance equations to ensure that the outputs are closer to a known color map, and this without hindering the quality of the generation. Our method leads to new guidance equations. We show in the color guidance context that, the scaling of the guidance should not decrease but remains high throughout the diffusion process. In a second contribution, our guidance is applied in a compression framework, we combine both semantic and general color information on the image to decode the images at low cost. We show that our method is effective at improving fidelity and realism of compressed images at extremely low bit rates, when compared to other classical or more semantic oriented approaches.

Tom Bachard, Thomas Maugey

Semantic compression, a compression scheme where the distortion metric, typically MSE, is replaced with semantic fidelity metrics, tends to become more and more popular. Most recent semantic compression schemes rely on the foundation model CLIP. In this work, we extend such a scheme to image collection compression, where inter-item redundancy is taken into account during the coding phase. For that purpose, we first show that CLIP's latent space allows for easy semantic additions and subtractions. From this property, we define a dictionary-based multi-item codec that outperforms state-of-the-art generative codec in terms of compression rate, around $10^{-5}$ BPP per image, while not sacrificing semantic fidelity. We also show that the learned dictionary is of a semantic nature and works as a semantic projector for the semantic content of images.

Navid Mahmoudian Bidgoli, Roberto G. de A. Azevedo, Thomas Maugey et al.

State-of-the-art 2D image compression schemes rely on the power of convolutional neural networks (CNNs). Although CNNs offer promising perspectives for 2D image compression, extending such models to omnidirectional images is not straightforward. First, omnidirectional images have specific spatial and statistical properties that can not be fully captured by current CNN models. Second, basic mathematical operations composing a CNN architecture, e.g., translation and sampling, are not well-defined on the sphere. In this paper, we study the learning of representation models for omnidirectional images and propose to use the properties of HEALPix uniform sampling of the sphere to redefine the mathematical tools used in deep learning models for omnidirectional images. In particular, we: i) propose the definition of a new convolution operation on the sphere that keeps the high expressiveness and the low complexity of a classical 2D convolution; ii) adapt standard CNN techniques such as stride, iterative aggregation, and pixel shuffling to the spherical domain; and then iii) apply our new framework to the task of omnidirectional image compression. Our experiments show that our proposed on-the-sphere solution leads to a better compression gain that can save 13.7% of the bit rate compared to similar learned models applied to equirectangular images. Also, compared to learning models based on graph convolutional networks, our solution supports more expressive filters that can preserve high frequencies and provide a better perceptual quality of the compressed images. Such results demonstrate the efficiency of the proposed framework, which opens new research venues for other omnidirectional vision tasks to be effectively implemented on the sphere manifold.

Navid Mahmoudian Bidgoli, Thomas Maugey, Aline Roumy

In this paper, we propose a new interactive compression scheme for omnidirectional images. This requires two characteristics: efficient compression of data, to lower the storage cost, and random access ability to extract part of the compressed stream requested by the user (for reducing the transmission rate). For efficient compression, data needs to be predicted by a series of references that have been pre-defined and compressed. This contrasts with the spirit of random accessibility. We propose a solution for this problem based on incremental codes implemented by rate-adaptive channel codes. This scheme encodes the image while adapting to any user request and leads to an efficient coding that is flexible in extracting data depending on the available information at the decoder. Therefore, only the information that is needed to be displayed at the user's side is transmitted during the user's request, as if the request was already known at the encoder. The experimental results demonstrate that our coder obtains a better transmission rate than the state-of-the-art tile-based methods at a small cost in storage. Moreover, the transmission rate grows gradually with the size of the request and avoids a staircase effect, which shows the perfect suitability of our coder for interactive transmission.

Mira Rizkallah, Xin Su, Thomas Maugey et al.

The paper addresses the problem of energy compaction of dense 4D light fields by designing geometry-aware local graph-based transforms. Local graphs are constructed on super-rays that can be seen as a grouping of spatially and geometry-dependent angularly correlated pixels. Both non separable and separable transforms are considered. Despite the local support of limited size defined by the super-rays, the Laplacian matrix of the non separable graph remains of high dimension and its diagonalization to compute the transform eigen vectors remains computationally expensive. To solve this problem, we then perform the local spatio-angular transform in a separable manner. We show that when the shape of corresponding super-pixels in the different views is not isometric, the basis functions of the spatial transforms are not coherent, resulting in decreased correlation between spatial transform coefficients. We hence propose a novel transform optimization method that aims at preserving angular correlation even when the shapes of the super-pixels are not isometric. Experimental results show the benefit of the approach in terms of energy compaction. A coding scheme is also described to assess the rate-distortion perfomances of the proposed transforms and is compared to state of the art encoders namely HEVC and JPEG Pleno VM 1.1.

Mira Rizkallah, Thomas Maugey, Christine Guillemot

Graph-based transforms have been shown to be powerful tools in terms of image energy compaction. However, when the support increases to best capture signal dependencies, the computation of the basis functions becomes rapidly untractable. This problem is in particular compelling for high dimensional imaging data such as light fields. The use of local transforms with limited supports is a way to cope with this computational difficulty. Unfortunately, the locality of the support may not allow us to fully exploit long term signal dependencies present in both the spatial and angular dimensions in the case of light fields. This paper describes sampling and prediction schemes with local graph-based transforms enabling to efficiently compact the signal energy and exploit dependencies beyond the local graph support. The proposed approach is investigated and is shown to be very efficient in the context of spatio-angular transforms for quasi-lossless compression of light fields.

Rui Ma, Thomas Maugey, Pascal Frossard

In contrary to traditional media streaming services where a unique media content is delivered to different users, interactive multiview navigation applications enable users to choose their own viewpoints and freely navigate in a 3-D scene. The interactivity brings new challenges in addition to the classical rate-distortion trade-off, which considers only the compression performance and viewing quality. On the one hand, interactivity necessitates sufficient viewpoints for richer navigation; on the other hand, it requires to provide low bandwidth and delay costs for smooth navigation during view transitions. In this paper, we formally describe the novel trade-offs posed by the navigation interactivity and classical rate-distortion criterion. Based on an original formulation, we look for the optimal design of the data representation by introducing novel rate and distortion models and practical solving algorithms. Experiments show that the proposed data representation method outperforms the baseline solution by providing lower resource consumptions and higher visual quality in all navigation configurations, which certainly confirms the potential of the proposed data representation in practical interactive navigation systems.

Ana De Abreu, Laura Toni, Nikolaos Thomos et al.

We consider an interactive multiview video streaming (IMVS) system where clients select their preferred viewpoint in a given navigation window. To provide high quality IMVS, many high quality views should be transmitted to the clients. However, this is not always possible due to the limited and heterogeneous capabilities of the clients. In this paper, we propose a novel adaptive IMVS solution based on a layered multiview representation where camera views are organized into layered subsets to match the different clients constraints. We formulate an optimization problem for the joint selection of the views subsets and their encoding rates. Then, we propose an optimal and a reduced computational complexity greedy algorithms, both based on dynamic-programming. Simulation results show the good performance of our novel algorithms compared to a baseline algorithm, proving that an effective IMVS adaptive solution should consider the scene content and the client capabilities and their preferences in navigation.

Laura Toni, Thomas Maugey, Pascal Frossard

In multiview video systems, multiple cameras generally acquire the same scene from different perspectives, such that users have the possibility to select their preferred viewpoint. This results in large amounts of highly redundant data, which needs to be properly handled during encoding and transmission over resource-constrained channels. In this work, we study coding and transmission strategies in multicamera systems, where correlated sources send data through a bottleneck channel to a central server, which eventually transmits views to different interactive users. We propose a dynamic correlation-aware packet scheduling optimization under delay, bandwidth, and interactivity constraints. The optimization relies both on a novel rate-distortion model, which captures the importance of each view in the 3D scene reconstruction, and on an objective function that optimizes resources based on a client navigation model. The latter takes into account the distortion experienced by interactive clients as well as the distortion variations that might be observed by clients during multiview navigation. We solve the scheduling problem with a novel trellis-based solution, which permits to formally decompose the multivariate optimization problem thereby significantly reducing the computation complexity. Simulation results show the gain of the proposed algorithm compared to baseline scheduling policies. More in details, we show the gain offered by our dynamic scheduling policy compared to static camera allocation strategies and to schemes with constant coding strategies. Finally, we show that the best scheduling policy consistently adapts to the most likely user navigation path and that it minimizes distortion variations that can be very disturbing for users in traditional navigation systems.

Thomas Maugey, Antonio Ortega, Pascal Frossard

In this paper, we propose a new representation for multiview image sets. Our approach relies on graphs to describe geometry information in a compact and controllable way. The links of the graph connect pixels in different images and describe the proximity between pixels in the 3D space. These connections are dependent on the geometry of the scene and provide the right amount of information that is necessary for coding and reconstructing multiple views. This multiview image representation is very compact and adapts the transmitted geometry information as a function of the complexity of the prediction performed at the decoder side. To achieve this, our GBR adapts the accuracy of the geometry representation, in contrast with depth coding, which directly compresses with losses the original geometry signal. We present the principles of this graph-based representation (GBR) and we build a complete prototype coding scheme for multiview images. Experimental results demonstrate the potential of this new representation as compared to a depth-based approach. GBR can achieve a gain of 2 dB in reconstructed quality over depth-based schemes operating at similar rates.

Uday Takyar, Thomas Maugey, Pascal Frossard

Emerging applications in multiview streaming look for providing interactive navigation services to video players. The user can ask for information from any viewpoint with a minimum transmission delay. The purpose is to provide user with as much information as possible with least number of redundancies. The recent concept of navigation segment representation consists of regrouping a given number of viewpoints in one signal and transmitting them to the users according to their navigation path. The question of the best description strategy of these navigation segments is however still open. In this paper, we propose to represent and code navigation segments by a method that extends the recent layered depth image (LDI) format. It consists of describing the scene from a viewpoint with multiple images organized in layers corresponding to the different levels of occluded objects. The notion of extended LDI comes from the fact that the size of this image is adapted to take into account the sides of the scene also, in contrary to classical LDI. The obtained results show a significant rate-distortion gain compared to classical multiview compression approaches in navigation scenario.

Thomas Maugey, Ismael Daribo, Gene Cheung et al.

Enabling users to interactively navigate through different viewpoints of a static scene is a new interesting functionality in 3D streaming systems. While it opens exciting perspectives towards rich multimedia applications, it requires the design of novel representations and coding techniques in order to solve the new challenges imposed by interactive navigation. Interactivity clearly brings new design constraints: the encoder is unaware of the exact decoding process, while the decoder has to reconstruct information from incomplete subsets of data since the server can generally not transmit images for all possible viewpoints due to resource constrains. In this paper, we propose a novel multiview data representation that permits to satisfy bandwidth and storage constraints in an interactive multiview streaming system. In particular, we partition the multiview navigation domain into segments, each of which is described by a reference image and some auxiliary information. The auxiliary information enables the client to recreate any viewpoint in the navigation segment via view synthesis. The decoder is then able to navigate freely in the segment without further data request to the server; it requests additional data only when it moves to a different segment. We discuss the benefits of this novel representation in interactive navigation systems and further propose a method to optimize the partitioning of the navigation domain into independent segments, under bandwidth and storage constraints. Experimental results confirm the potential of the proposed representation; namely, our system leads to similar compression performance as classical inter-view coding, while it provides the high level of flexibility that is required for interactive streaming. Hence, our new framework represents a promising solution for 3D data representation in novel interactive multimedia services.

Thomas Maugey, Pascal Frossard

Multiview video with interactive and smooth view switching at the receiver is a challenging application with several issues in terms of effective use of storage and bandwidth resources, reactivity of the system, quality of the viewing experience and system complexity. The classical decoding system for generating virtual views first projects a reference or encoded frame to a given viewpoint and then fills in the holes due to potential occlusions. This last step still constitutes a complex operation with specific software or hardware at the receiver and requires a certain quantity of information from the neighboring frames for insuring consistency between the virtual images. In this work we propose a new approach that shifts most of the burden due to interactivity from the decoder to the encoder, by anticipating the navigation of the decoder and sending auxiliary information that guarantees temporal and interview consistency. This leads to an additional cost in terms of transmission rate and storage, which we minimize by using optimization techniques based on the user behavior modeling. We show by experiments that the proposed system represents a valid solution for interactive multiview systems with classical decoders.