Laura Toni

Clement Vignac, Nagham Osman, Laura Toni et al.

This work introduces MiDi, a novel diffusion model for jointly generating molecular graphs and their corresponding 3D arrangement of atoms. Unlike existing methods that rely on predefined rules to determine molecular bonds based on the 3D conformation, MiDi offers an end-to-end differentiable approach that streamlines the molecule generation process. Our experimental results demonstrate the effectiveness of this approach. On the challenging GEOM-DRUGS dataset, MiDi generates 92% of stable molecules, against 6% for the previous EDM model that uses interatomic distances for bond prediction, and 40% using EDM followed by an algorithm that directly optimize bond orders for validity. Our code is available at github.com/cvignac/MiDi.

Eduardo Pignatelli, Johan Ferret, Tim Rockäschel et al.

The temporal credit assignment problem is a central challenge in Reinforcement Learning (RL), concerned with attributing the appropriate influence to each actions in a trajectory for their ability to achieve a goal. However, when feedback is delayed and sparse, the learning signal is poor, and action evaluation becomes harder. Canonical solutions, such as reward shaping and options, require extensive domain knowledge and manual intervention, limiting their scalability and applicability. In this work, we lay the foundations for Credit Assignment with Language Models (CALM), a novel approach that leverages Large Language Models (LLMs) to automate credit assignment via reward shaping and options discovery. CALM uses LLMs to decompose a task into elementary subgoals and assess the achievement of these subgoals in state-action transitions. Every time an option terminates, a subgoal is achieved, and CALM provides an auxiliary reward. This additional reward signal can enhance the learning process when the task reward is sparse and delayed without the need for human-designed rewards. We provide a preliminary evaluation of CALM using a dataset of human-annotated demonstrations from MiniHack, suggesting that LLMs can be effective in assigning credit in zero-shot settings, without examples or LLM fine-tuning. Our preliminary results indicate that the knowledge of LLMs is a promising prior for credit assignment in RL, facilitating the transfer of human knowledge into value functions.

Gholamali Aminian, Saeed Masiha, Laura Toni et al.

Generalization error bounds are essential for comprehending how well machine learning models work. In this work, we suggest a novel method, i.e., the Auxiliary Distribution Method, that leads to new upper bounds on expected generalization errors that are appropriate for supervised learning scenarios. We show that our general upper bounds can be specialized under some conditions to new bounds involving the $α$-Jensen-Shannon, $α$-Rényi ($0< α< 1$) information between a random variable modeling the set of training samples and another random variable modeling the set of hypotheses. Our upper bounds based on $α$-Jensen-Shannon information are also finite. Additionally, we demonstrate how our auxiliary distribution method can be used to derive the upper bounds on excess risk of some learning algorithms in the supervised learning context {\blue and the generalization error under the distribution mismatch scenario in supervised learning algorithms, where the distribution mismatch is modeled as $α$-Jensen-Shannon or $α$-Rényi divergence between the distribution of test and training data samples distributions.} We also outline the conditions for which our proposed upper bounds might be tighter than other earlier upper bounds.

Laura Toni, Pascal Frossard

We propose Grab-UCB, a graph-kernel multi-arms bandit algorithm to learn online the optimal source placement in large scale networks, such that the reward obtained from a priori unknown network processes is maximized. The uncertainty calls for online learning, which suffers however from the curse of dimensionality. To achieve sample efficiency, we describe the network processes with an adaptive graph dictionary model, which typically leads to sparse spectral representations. This enables a data-efficient learning framework, whose learning rate scales with the dimension of the spectral representation model instead of the one of the network. We then propose Grab-UCB, an online sequential decision strategy that learns the parameters of the spectral representation while optimizing the action strategy. We derive the performance guarantees that depend on network parameters, which further influence the learning curve of the sequential decision strategy We introduce a computationally simplified solving method, Grab-arm-Light, an algorithm that walks along the edges of the polytope representing the objective function. Simulations results show that the proposed online learning algorithm outperforms baseline offline methods that typically separate the learning phase from the testing one. The results confirm the theoretical findings, and further highlight the gain of the proposed online learning strategy in terms of cumulative regret, sample efficiency and computational complexity.

Gholamali Aminian, Armin Behnamnia, Roberto Vega et al.

Off-policy learning methods are intended to learn a policy from logged data, which includes context, action, and feedback (cost or reward) for each sample point. In this work, we build on the counterfactual risk minimization framework, which also assumes access to propensity scores. We propose learning methods for problems where feedback is missing for some samples, so there are samples with feedback and samples missing-feedback in the logged data. We refer to this type of learning as semi-supervised batch learning from logged data, which arises in a wide range of application domains. We derive a novel upper bound for the true risk under the inverse propensity score estimator to address this kind of learning problem. Using this bound, we propose a regularized semi-supervised batch learning method with logged data where the regularization term is feedback-independent and, as a result, can be evaluated using the logged missing-feedback data. Consequently, even though feedback is only present for some samples, a learning policy can be learned by leveraging the missing-feedback samples. The results of experiments derived from benchmark datasets indicate that these algorithms achieve policies with better performance in comparison with logging policies.

Eduardo Pignatelli, Jarek Liesen, Robert Tjarko Lange et al.

As Deep Reinforcement Learning (Deep RL) research moves towards solving large-scale worlds, efficient environment simulations become crucial for rapid experimentation. However, most existing environments struggle to scale to high throughput, setting back meaningful progress. Interactions are typically computed on the CPU, limiting training speed and throughput, due to slower computation and communication overhead when distributing the task across multiple machines. Ultimately, Deep RL training is CPU-bound, and developing batched, fast, and scalable environments has become a frontier for progress. Among the most used Reinforcement Learning (RL) environments, MiniGrid is at the foundation of several studies on exploration, curriculum learning, representation learning, diversity, meta-learning, credit assignment, and language-conditioned RL, and still suffers from the limitations described above. In this work, we introduce NAVIX, a re-implementation of MiniGrid in JAX. NAVIX achieves over 200 000x speed improvements in batch mode, supporting up to 2048 agents in parallel on a single Nvidia A100 80 GB. This reduces experiment times from one week to 15 minutes, promoting faster design iterations and more scalable RL model development.

Tommaso Giorgi, Pierriccardo Olivieri, Keyue Jiang et al.

Learning compact state representations in Markov Decision Processes (MDPs) has proven crucial for addressing the curse of dimensionality in large-scale reinforcement learning (RL) problems. Existing principled approaches leverage structural priors on the MDP by constructing state representations as linear combinations of the state-graph Laplacian eigenvectors. When the transition graph is unknown or the state space is prohibitively large, the graph spectral features can be estimated directly via sample trajectories. In this work, we prove an upper bound on the approximation error of linear value function approximation under the learned spectral features. We show how this error scales with the algebraic connectivity of the state-graph, grounding the approximation quality in the topological structure of the MDP. We further bound the error introduced by the eigenvector estimation itself, leading to an end-to-end error decomposition across the representation learning pipeline. Additionally, our expression of the Laplacian operator for the RL setting, although equivalent to existing ones, prevents some common misunderstandings, of which we show some examples from the literature. Our results hold for general (non-uniform) policies without any assumptions on the symmetry of the induced transition kernel. We validate our theoretical findings with numerical simulations on gridworld environments.

Alexandru Cioba, Aya Kayal, Laura Toni et al.

In many real-world reinforcement learning (RL) problems, the environment exhibits inherent symmetries that can be exploited to improve learning efficiency. This paper develops a theoretical and algorithmic framework for incorporating known group symmetries into kernel-based RL. We propose a symmetry-aware variant of optimistic least-squares value iteration (LSVI), which leverages invariant kernels to encode invariance in both rewards and transition dynamics. Our analysis establishes new bounds on the maximum information gain and covering numbers for invariant RKHSs, explicitly quantifying the sample efficiency gains from symmetry. Empirical results on a customized Frozen Lake environment and a 2D placement design problem confirm the theoretical improvements, demonstrating that symmetry-aware RL achieves significantly better performance than their standard kernel counterparts. These findings highlight the value of structural priors in designing more sample-efficient reinforcement learning algorithms.

Pedro Gomes, Silvia Rossi, Laura Toni

This paper focuses on motion prediction for point cloud sequences in the challenging case of deformable 3D objects, such as human body motion. First, we investigate the challenges caused by deformable shapes and complex motions present in this type of representation, with the ultimate goal of understanding the technical limitations of state-of-the-art models. From this understanding, we propose an improved architecture for point cloud prediction of deformable 3D objects. Specifically, to handle deformable shapes, we propose a graph-based approach that learns and exploits the spatial structure of point clouds to extract more representative features. Then we propose a module able to combine the learned features in an adaptative manner according to the point cloud movements. The proposed adaptative module controls the composition of local and global motions for each point, enabling the network to model complex motions in deformable 3D objects more effectively. We tested the proposed method on the following datasets: MNIST moving digits, the Mixamo human bodies motions, JPEG and CWIPC-SXR real-world dynamic bodies. Simulation results demonstrate that our method outperforms the current baseline methods given its improved ability to model complex movements as well as preserve point cloud shape. Furthermore, we demonstrate the generalizability of the proposed framework for dynamic feature learning, by testing the framework for action recognition on the MSRAction3D dataset and achieving results on-par with state-of-the-art methods

Aya Kayal, Eduardo Pignatelli, Laura Toni

One of the open challenges in Reinforcement Learning is the hard exploration problem in sparse reward environments. Various types of intrinsic rewards have been proposed to address this challenge by pushing towards diversity. This diversity might be imposed at different levels, favouring the agent to explore different states, policies or behaviours (State, Policy and Skill level diversity, respectively). However, the impact of diversity on the agent's behaviour remains unclear. In this work, we aim to fill this gap by studying the effect of different levels of diversity imposed by intrinsic rewards on the exploration patterns of RL agents. We select four intrinsic rewards (State Count, Intrinsic Curiosity Module (ICM), Maximum Entropy, and Diversity is all you need (DIAYN)), each pushing for a different diversity level. We conduct an empirical study on MiniGrid environment to compare their impact on exploration considering various metrics related to the agent's exploration, namely: episodic return, observation coverage, agent's position coverage, policy entropy, and timeframes to reach the sparse reward. The main outcome of the study is that State Count leads to the best exploration performance in the case of low-dimensional observations. However, in the case of RGB observations, the performance of State Count is highly degraded mostly due to representation learning challenges. Conversely, Maximum Entropy is less impacted, resulting in a more robust exploration, despite being not always optimal. Lastly, our empirical study revealed that learning diverse skills with DIAYN, often linked to improved robustness and generalisation, does not promote exploration in MiniGrid environments. This is because: i) learning the skill space itself can be challenging, and ii) exploration within the skill space prioritises differentiating between behaviours rather than achieving uniform state visitation.

Aya Kayal, Sattar Vakili, Laura Toni et al.

Bayesian optimization (BO) with preference-based feedback has recently garnered significant attention due to its emerging applications. We refer to this problem as Bayesian Optimization from Human Feedback (BOHF), which differs from conventional BO by learning the best actions from a reduced feedback model, where only the preference between two actions is revealed to the learner at each time step. The objective is to identify the best action using a limited number of preference queries, typically obtained through costly human feedback. Existing work, which adopts the Bradley-Terry-Luce (BTL) feedback model, provides regret bounds for the performance of several algorithms. In this work, within the same framework we develop tighter performance guarantees. Specifically, we derive regret bounds of $\tilde{\mathcal{O}}(\sqrt{Γ(T)T})$, where $Γ(T)$ represents the maximum information gain$\unicode{x2014}$a kernel-specific complexity term$\unicode{x2014}$and $T$ is the number of queries. Our results significantly improve upon existing bounds. Notably, for common kernels, we show that the order-optimal sample complexities of conventional BO$\unicode{x2014}$achieved with richer feedback models$\unicode{x2014}$are recovered. In other words, the same number of preferential samples as scalar-valued samples is sufficient to find a nearly optimal solution.

Aya Kayal, Sattar Vakili, Laura Toni et al.

Reinforcement Learning (RL) problems are being considered under increasingly more complex structures. While tabular and linear models have been thoroughly explored, the analytical study of RL under nonlinear function approximation, especially kernel-based models, has recently gained traction for their strong representational capacity and theoretical tractability. In this context, we examine the question of statistical efficiency in kernel-based RL within the reward-free RL framework, specifically asking: how many samples are required to design a near-optimal policy? Existing work addresses this question under restrictive assumptions about the class of kernel functions. We first explore this question by assuming a generative model, then relax this assumption at the cost of increasing the sample complexity by a factor of H, the length of the episode. We tackle this fundamental problem using a broad class of kernels and a simpler algorithm compared to prior work. Our approach derives new confidence intervals for kernel ridge regression, specific to our RL setting, which may be of broader applicability. We further validate our theoretical findings through simulations.

Gholamali Aminian, Mahed Abroshan, Mohammad Mahdi Khalili et al.

A common assumption in semi-supervised learning is that the labeled, unlabeled, and test data are drawn from the same distribution. However, this assumption is not satisfied in many applications. In many scenarios, the data is collected sequentially (e.g., healthcare) and the distribution of the data may change over time often exhibiting so-called covariate shifts. In this paper, we propose an approach for semi-supervised learning algorithms that is capable of addressing this issue. Our framework also recovers some popular methods, including entropy minimization and pseudo-labeling. We provide new information-theoretical based generalization error upper bounds inspired by our novel framework. Our bounds are applicable to both general semi-supervised learning and the covariate-shift scenario. Finally, we show numerically that our method outperforms previous approaches proposed for semi-supervised learning under the covariate shift.

Silvia Rossi, Irene Viola, Laura Toni et al.

Immersive reality technologies, such as Virtual and Augmented Reality, have ushered a new era of user-centric systems, in which every aspect of the coding--delivery--rendering chain is tailored to the interaction of the users. Understanding the actual interactivity and behaviour of the users is still an open challenge and a key step to enabling such a user-centric system. Our main goal is to extend the applicability of existing behavioural methodologies for studying user navigation in the case of 6 Degree-of-Freedom (DoF). Specifically, we first compare the navigation in 6-DoF with its 3-DoF counterpart highlighting the main differences and novelties. Then, we define new metrics aimed at better modelling behavioural similarities between users in a 6-DoF system. We validate and test our solutions on real navigation paths of users interacting with dynamic volumetric media in 6-DoF Virtual Reality conditions. Our results show that metrics that consider both user position and viewing direction better perform in detecting user similarity while navigating in a 6-DoF system. Having easy-to-use but robust metrics that underpin multiple tools and answer the question ``how do we detect if two users look at the same content?" open the gate to new solutions for a user-centric system.

Yuheng Bu, Gholamali Aminian, Laura Toni et al.

We provide an information-theoretic analysis of the generalization ability of Gibbs-based transfer learning algorithms by focusing on two popular transfer learning approaches, $α$-weighted-ERM and two-stage-ERM. Our key result is an exact characterization of the generalization behaviour using the conditional symmetrized KL information between the output hypothesis and the target training samples given the source samples. Our results can also be applied to provide novel distribution-free generalization error upper bounds on these two aforementioned Gibbs algorithms. Our approach is versatile, as it also characterizes the generalization errors and excess risks of these two Gibbs algorithms in the asymptotic regime, where they converge to the $α$-weighted-ERM and two-stage-ERM, respectively. Based on our theoretical results, we show that the benefits of transfer learning can be viewed as a bias-variance trade-off, with the bias induced by the source distribution and the variance induced by the lack of target samples. We believe this viewpoint can guide the choice of transfer learning algorithms in practice.

Gholamali Aminian, Yuheng Bu, Laura Toni et al.

Bounding the generalization error of a supervised learning algorithm is one of the most important problems in learning theory, and various approaches have been developed. However, existing bounds are often loose and lack of guarantees. As a result, they may fail to characterize the exact generalization ability of a learning algorithm. Our main contribution is an exact characterization of the expected generalization error of the well-known Gibbs algorithm in terms of symmetrized KL information between the input training samples and the output hypothesis. Such a result can be applied to tighten existing expected generalization error bound. Our analysis provides more insight on the fundamental role the symmetrized KL information plays in controlling the generalization error of the Gibbs algorithm.

Pedro Gomes, Silvia Rossi, Laura Toni

In this paper, we propose an end-to-end learning network to predict future frames in a point cloud sequence. As main novelty, an initial layer learns topological information of point clouds as geometric features, to form representative spatio-temporal neighborhoods. This module is followed by multiple Graph-RNN cells. Each cell learns points dynamics (i.e., RNN states) by processing each point jointly with the spatio-temporal neighbouring points. We tested the network performance with a MINST dataset of moving digits, a synthetic human bodies motions and JPEG dynamic bodies datasets. Simulation results demonstrate that our method outperforms baseline ones that neglect geometry features information.

Gholamali Aminian, Laura Toni, Miguel R. D. Rodrigues

Generalization error bounds are critical to understanding the performance of machine learning models. In this work, building upon a new bound of the expected value of an arbitrary function of the population and empirical risk of a learning algorithm, we offer a more refined analysis of the generalization behaviour of a machine learning models based on a characterization of (bounds) to their generalization error moments. We discuss how the proposed bounds -- which also encompass new bounds to the expected generalization error -- relate to existing bounds in the literature. We also discuss how the proposed generalization error moment bounds can be used to construct new generalization error high-probability bounds.

Gholamali Aminian, Laura Toni, Miguel R. D. Rodrigues

Generalization error bounds are critical to understanding the performance of machine learning models. In this work, we propose a new information-theoretic based generalization error upper bound applicable to supervised learning scenarios. We show that our general bound can specialize in various previous bounds. We also show that our general bound can be specialized under some conditions to a new bound involving the Jensen-Shannon information between a random variable modelling the set of training samples and another random variable modelling the hypothesis. We also prove that our bound can be tighter than mutual information-based bounds under some conditions.

Xiaowen Dong, Dorina Thanou, Laura Toni et al.

The effective representation, processing, analysis, and visualization of large-scale structured data, especially those related to complex domains such as networks and graphs, are one of the key questions in modern machine learning. Graph signal processing (GSP), a vibrant branch of signal processing models and algorithms that aims at handling data supported on graphs, opens new paths of research to address this challenge. In this article, we review a few important contributions made by GSP concepts and tools, such as graph filters and transforms, to the development of novel machine learning algorithms. In particular, our discussion focuses on the following three aspects: exploiting data structure and relational priors, improving data and computational efficiency, and enhancing model interpretability. Furthermore, we provide new perspectives on future development of GSP techniques that may serve as a bridge between applied mathematics and signal processing on one side, and machine learning and network science on the other. Cross-fertilization across these different disciplines may help unlock the numerous challenges of complex data analysis in the modern age.

Kaige Yang, Laura Toni

Upper Confidence Bound (UCB) is arguably the most commonly used method for linear multi-arm bandit problems. While conceptually and computationally simple, this method highly relies on the confidence bounds, failing to strike the optimal exploration-exploitation if these bounds are not properly set. In the literature, confidence bounds are typically derived from concentration inequalities based on assumptions on the reward distribution, e.g., sub-Gaussianity. The validity of these assumptions however is unknown in practice. In this work, we aim at learning the confidence bound in a data-driven fashion, making it adaptive to the actual problem structure. Specifically, noting that existing UCB-typed algorithms are not differentiable with respect to confidence bound, we first propose a novel differentiable linear bandit algorithm. Then, we introduce a gradient estimator, which allows the confidence bound to be learned via gradient ascent. Theoretically, we show that the proposed algorithm achieves a $\tilde{\mathcal{O}}(\hatβ\sqrt{dT})$ upper bound of $T$-round regret, where $d$ is the dimension of arm features and $\hatβ$ is the learned size of confidence bound. Empirical results show that $\hatβ$ is significantly smaller than its theoretical upper bound and proposed algorithms outperforms baseline ones on both simulated and real-world datasets.

Sephora Madjiheurem, Laura Toni

A major challenge in reinforcement learning (RL) is the design of agents that are able to generalize across tasks that share common dynamics. A viable solution is meta-reinforcement learning, which identifies common structures among past tasks to be then generalized to new tasks (meta-test). In meta-training, the RL agent learns state representations that encode prior information from a set of tasks, used to generalize the value function approximation. This has been proposed in the literature as successor representation approximators. While promising, these methods do not generalize well across optimal policies, leading to sampling-inefficiency during meta-test phases. In this paper, we propose state2vec, an efficient and low-complexity framework for learning successor features which (i) generalize across policies, (ii) ensure sample-efficiency during meta-test. We extend the well known node2vec framework to learn state embeddings that account for the discounted future state transitions in RL. The proposed off-policy state2vec captures the geometry of the underlying state space, making good basis functions for linear value function approximation.

Kaige Yang, Xiaowen Dong, Laura Toni

We consider a stochastic linear bandit problem with multiple users, where the relationship between users is captured by an underlying graph and user preferences are represented as smooth signals on the graph. We introduce a novel bandit algorithm where the smoothness prior is imposed via the random-walk graph Laplacian, which leads to a single-user cumulative regret scaling as $\tilde{\mathcal{O}}(Ψd \sqrt{T})$ with time horizon $T$, feature dimensionality $d$, and the scalar parameter $Ψ\in (0,1)$ that depends on the graph connectivity. This is an improvement over $\tilde{\mathcal{O}}(d \sqrt{T})$ in \algo{LinUCB}~\Ccite{li2010contextual}, where user relationship is not taken into account. In terms of network regret (sum of cumulative regret over $n$ users), the proposed algorithm leads to a scaling as $\tilde{\mathcal{O}}(Ψd\sqrt{nT})$, which is a significant improvement over $\tilde{\mathcal{O}}(nd\sqrt{T})$ in the state-of-the-art algorithm \algo{Gob.Lin} \Ccite{cesa2013gang}. To improve scalability, we further propose a simplified algorithm with a linear computational complexity with respect to the number of users, while maintaining the same regret. Finally, we present a finite-time analysis on the proposed algorithms, and demonstrate their advantage in comparison with state-of-the-art graph-based bandit algorithms on both synthetic and real-world data.

Kaige Yang, Xiaowen Dong, Laura Toni

We provide a theoretical analysis of the representation learning problem aimed at learning the latent variables (design matrix) $Θ$ of observations $Y$ with the knowledge of the coefficient matrix $X$. The design matrix is learned under the assumption that the latent variables $Θ$ are smooth with respect to a (known) topological structure $\mathcal{G}$. To learn such latent variables, we study a graph Laplacian regularized estimator, which is the penalized least squares estimator with penalty term proportional to a Laplacian quadratic form. This type of estimators has recently received considerable attention due to its capability in incorporating underlying topological graph structure of variables into the learning process. While the estimation problem can be solved efficiently by state-of-the-art optimization techniques, its statistical consistency properties have been largely overlooked. In this work, we develop a non-asymptotic bound of estimation error under the classical statistical setting, where sample size is larger than the ambient dimension of the latent variables. This bound illustrates theoretically the impact of the alignment between the data and the graph structure as well as the graph spectrum on the estimation accuracy. It also provides theoretical evidence of the advantage, in terms of convergence rate, of the graph Laplacian regularized estimator over classical ones (that ignore the graph structure) in case of a smoothness prior. Finally, we provide empirical results of the estimation error to corroborate the theoretical analysis.

Sephora Madjiheurem, Laura Toni

In this work, we study value function approximation in reinforcement learning (RL) problems with high dimensional state or action spaces via a generalized version of representation policy iteration (RPI). We consider the limitations of proto-value functions (PVFs) at accurately approximating the value function in low dimensions and we highlight the importance of features learning for an improved low-dimensional value function approximation. Then, we adopt different representation learning algorithm on graphs to learn the basis functions that best represent the value function. We empirically show that node2vec, an algorithm for scalable feature learning in networks, and the Variational Graph Auto-Encoder constantly outperform the commonly used smooth proto-value functions in low-dimensional feature space.

Silvia Rossi, Francesca De Simone, Pascal Frossard et al.

In Virtual Reality (VR) applications, understanding how users explore the omnidirectional content is important to optimize content creation, to develop user-centric services, or even to detect disorders in medical applications. Clustering users based on their common navigation patterns is a first direction to understand users behaviour. However, classical clustering techniques fail in identifying these common paths, since they are usually focused on minimizing a simple distance metric. In this paper, we argue that minimizing the distance metric does not necessarily guarantee to identify users that experience similar navigation path in the VR domain. Therefore, we propose a graph-based method to identify clusters of users who are attending the same portion of the spherical content over time. The proposed solution takes into account the spherical geometry of the content and aims at clustering users based on the actual overlap of displayed content among users. Our method is tested on real VR user navigation patterns. Results show that our solution leads to clusters in which at least 85% of the content displayed by one user is shared among the other users belonging to the same cluster.

Kaige Yang, Laura Toni

In this work, we study recommendation systems modelled as contextual multi-armed bandit (MAB) problems. We propose a graph-based recommendation system that learns and exploits the geometry of the user space to create meaningful clusters in the user domain. This reduces the dimensionality of the recommendation problem while preserving the accuracy of MAB. We then study the effect of graph sparsity and clusters size on the MAB performance and provide exhaustive simulation results both in synthetic and in real-case datasets. Simulation results show improvements with respect to state-of-the-art MAB algorithms.

Xue Zhang, Laura Toni, Pascal Frossard et al.

Multiview applications endow final users with the possibility to freely navigate within 3D scenes with minimum-delay. A real feeling of scene navigation is enabled by transmitting multiple high-quality camera views, which can be used to synthesize additional virtual views to offer a smooth navigation. However, when network resources are limited, not all camera views can be sent at high quality. It is therefore important, yet challenging, to find the right tradeoff between coding artifacts (reducing the quality of camera views) and virtual synthesis artifacts (reducing the number of camera views sent to users). To this aim, we propose an optimal transmission strategy for interactive multiview HTTP adaptive streaming (HAS). We propose a problem formulation to select the optimal set of camera views that the client requests for downloading, such that the navigation quality experienced by the user is optimized while the bandwidth constraints are satisfied. We show that our optimization problem is NP-hard, and we therefore develop an optimal solution based on the dynamic programming algorithm with polynomial time complexity. To further simplify the deployment, we present a suboptimal greedy algorithm with effective performance and lower complexity. The proposed controller is evaluated in theoretical and realistic settings characterized by realistic network statistics estimation, buffer management and server-side representation optimization. Simulation results show significant improvement in terms of navigation quality compared with alternative baseline multiview adaptation logic solutions.

Stefano D'Aronco, Laura Toni, Pascal Frossard

HTTP adaptive streaming (HAS) has become the universal technology for video streaming over the Internet. Many HAS system designs aim at sharing the network bandwidth in a rate-fair manner. However, rate fairness is in general not equivalent to quality fairness as different video sequences might have different characteristics and resource requirements. In this work, we focus on this limitation and propose a novel controller for HAS clients that is able to reach quality fairness while preserving the main characteristics of HAS systems and with a limited support from the network devices. In particular, we adopt a price-based mechanism in order to build a controller that maximizes the aggregate video quality for a set of HAS clients that share a common bottleneck. When network resources are scarce, the clients with simple video sequences reduce the requested bitrate in favor of users that subscribe to more complex video sequences, leading to a more efficient network usage. The proposed controller has been implemented in a network simulator, and the simulation results demonstrate its ability to share the available bandwidth among the HAS users in a quality-fair manner.

Laura Toni, Pascal Frossard

Interactive multi-view video streaming (IMVS) services permit to remotely immerse within a 3D scene. This is possible by transmitting a set of reference camera views (anchor views), which are used by the clients to freely navigate in the scene and possibly synthesize additional viewpoints of interest. From a networking perspective, the big challenge in IMVS systems is to deliver to each client the best set of anchor views that maximizes the navigation quality, minimizes the view-switching delay and yet satisfies the network constraints. Integrating adaptive streaming solutions in free-viewpoint systems offers a promising solution to deploy IMVS in large and heterogeneous scenarios, as long as the multi-view video representations on the server are properly selected. We therefore propose to optimize the multi-view data at the server by minimizing the overall resource requirements, yet offering a good navigation quality to the different users. We propose a video representation set optimization for multiview adaptive streaming systems and we show that it is NP-hard. We therefore introduce the concept of multi-view navigation segment that permits to cast the video representation set selection as an integer linear programming problem with a bounded computational complexity. We then show that the proposed solution reduces the computational complexity while preserving optimality in most of the 3D scenes. We then provide simulation results for different classes of users and show the gain offered by an optimal multi-view video representation selection compared to recommended representation sets (e.g., Netflix and Apple ones) or to a baseline representation selection algorithm where the encoding parameters are decided a priori for all the views.

Laura Toni, Gene Cheung, Pascal Frossard

To enable Interactive multiview video systems with a minimum view-switching delay, multiple camera views are sent to the users, which are used as reference images to synthesize additional virtual views via depth-image-based rendering. In practice, bandwidth constraints may however restrict the number of reference views sent to clients per time unit, which may in turn limit the quality of the synthesized viewpoints. We argue that the reference view selection should ideally be performed close to the users, and we study the problem of in-network reference view synthesis such that the navigation quality is maximized at the clients. We consider a distributed cloud network architecture where data stored in a main cloud is delivered to end users with the help of cloudlets, i.e., resource-rich proxies close to the users. In order to satisfy last-hop bandwidth constraints from the cloudlet to the users, a cloudlet re-samples viewpoints of the 3D scene into a discrete set of views (combination of received camera views and virtual views synthesized) to be used as reference for the synthesis of additional virtual views at the client. This in-network synthesis leads to better viewpoint sampling given a bandwidth constraint compared to simple selection of camera views, but it may however carry a distortion penalty in the cloudlet-synthesized reference views. We therefore cast a new reference view selection problem where the best subset of views is defined as the one minimizing the distortion over a view navigation window defined by the user under some transmission bandwidth constraints. We show that the view selection problem is NP-hard, and propose an effective polynomial time algorithm using dynamic programming to solve the optimization problem. Simulation results finally confirm the performance gain offered by virtual view synthesis in the network.

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.

Laura Toni, Ramon Aparicio-Pardo, Karine Pires et al.

Adaptive streaming addresses the increasing and heterogenous demand of multimedia content over the Internet by offering several encoded versions for each video sequence. Each version (or representation) has a different resolution and bit rate, aimed at a specific set of users, like TV or mobile phone clients. While most existing works on adaptive streaming deal with effective playout-control strategies at the client side, we take in this paper a providers' perspective and propose solutions to improve user satisfaction by optimizing the encoding rates of the video sequences. We formulate an integer linear program that maximizes users' average satisfaction, taking into account the network dynamics, the video content information, and the user population characteristics. The solution of the optimization is a set of encoding parameters that permit to create different streams to robustly satisfy users' requests over time. We simulate multiple adaptive streaming sessions characterized by realistic network connections models, where the proposed solution outperforms commonly used vendor recommendations, in terms of user satisfaction but also in terms of fairness and outage probability. The simulation results further show that video content information as well as network constraints and users' statistics play a crucial role in selecting proper encoding parameters to provide fairness a mong users and to reduce network resource usage. We finally propose a few practical guidelines that can be used to choose the encoding parameters based on the user base characteristics, the network capacity and the type of video content.