A person downloading a pre-trained model from the web should be aware of its biases. Existing approaches for bias identification rely on datasets containing labels for the task of interest, something that a non-expert may not have access to, or may not have the necessary resources to collect: this greatly limits the number of tasks where model biases can be identified. In this work, we present Classifier-to-Bias (C2B), the first bias discovery framework that works without access to any labeled data: it only relies on a textual description of the classification task to identify biases in the target classification model. This description is fed to a large language model to generate bias proposals and corresponding captions depicting biases together with task-specific target labels. A retrieval model collects images for those captions, which are then used to assess the accuracy of the model w.r.t. the given biases. C2B is training-free, does not require any annotations, has no constraints on the list of biases, and can be applied to any pre-trained model on any classification task. Experiments on two publicly available datasets show that C2B discovers biases beyond those of the original datasets and outperforms a recent state-of-the-art bias detection baseline that relies on task-specific annotations, being a promising first step toward addressing task-agnostic unsupervised bias detection.

Prediction of head movements in immersive media is key to designing efficient streaming systems able to focus the bandwidth budget on visible areas of the content. However, most of the numerous proposals made to predict user head motion in 360° images and videos do not explicitly consider a prominent characteristic of the head motion data: its intrinsic uncertainty. In this article, we present an approach to generate multiple plausible futures of head motion in 360° videos, given a common past trajectory. To our knowledge, this is the first work that considers the problem of multiple head motion prediction for 360° video streaming. We introduce our discrete variational multiple sequence (DVMS) learning framework, which builds on deep latent variable models. We design a training procedure to obtain a flexible, lightweight stochastic prediction model compatible with sequence-to-sequence neural architectures. Experimental results on four different datasets show that DVMS outperforms competitors adapted from the self-driving domain by up to 41% on prediction horizons up to 5 s, at lower computational and memory costs. To understand how the learned features account for the motion uncertainty, we analyze the structure of the learned latent space and connect it with the physical properties of the trajectories. We also introduce a method to estimate the likelihood of each generated trajectory, enabling the integration of DVMS in a streaming system. We hence deploy an extensive evaluation of the interest of our DVMS proposal for a streaming system. To do so, we first introduce a new Python-based 360° streaming simulator that we make available to the community. On real-world user, video, and networking data, we show that predicting multiple trajectories yields higher fairness between the traces, the gains for 20–30% of the users reaching up to 10% in visual quality for the best number K of trajectories to generate.

Adaptive bitrate (ABR) algorithms are used in streaming media to adjust video or audio quality based on the viewer's network conditions to provide a smooth playback experience. With the rise of virtual reality (VR) headsets, 360° video streaming is growing rapidly and requires efficient ABR strategies to also adapt the video quality to the user's head position. However, research in this field is often difficult to compare due to a lack of reproducible simulations. To address this problem, we provide SMART360, a 360° streaming simulation environment to compare motion prediction and adaptive bitrates strategies. We provide sample inputs and baseline algorithms along with the simulator, as well as examples of results and visualizations that can be obtained with SMART360. The code and data are made publicly available.

While immersive media have been shown to generate more intense emotions, saliency information has been shown to be a key component for the assessment of their quality, owing to the various portions of the sphere (viewports) a user can attend. In this article, we investigate the tri-partite connection between user attention, user emotion and visual content in immersive environments. To do so, we present a new dataset enabling the analysis of different types of saliency, both low-level and high-level, in connection with the user’s state in 360◦videos. Head and gaze movements are recorded along with self-reports and continuous physiological measurements of emotions. We then study how the accuracy of saliency estimators in predicting user attention depends on user-reported and physiologically-sensed emotional perceptions. Our results show that high-level saliency better predicts user attention for higher levels of arousal. We discuss how this work serves as a first step to understand and predict user attention and intents in immersive interactive environments.

From a user perspective, immersive content can elicit more intense emotions than flat-screen presentations. From a system perspective, efficient storage and distribution remain challenging, and must consider user attention. Understanding the connection between user attention, user emotions and immersive content is therefore key. In this article, we present a new dataset, PEM360 of user head movements and gaze recordings in 360° videos, along with self-reported emotional ratings of valence and arousal, and continuous physiological measurement of electrodermal activity and heart rate. The stimuli are selected to enable the spatiotemporal analysis of the connection between content, user motion and emotion. We describe and provide a set of software tools to process the various data modalities, and introduce a joint instantaneous visualization of user attention and emotion we name Emotional maps. We exemplify new types of analyses the PEM360 dataset can enable. The entire data and code are made available in a reproducible framework.

This paper contains the research proposal of Quentin Guimard that was presented at the MMSys 2022 doctoral symposium.

Prediction of head movements in immersive media is key to design efficient streaming systems able to focus the bandwidth budget on visible areas of the content. Numerous proposals have therefore been made in the recent years to predict 360° images and videos. However, the performance of these models is limited by a main characteristic of the head motion data: its intrinsic uncertainty. In this article, we present an approach to generate multiple plausible futures of head motion in 360° videos, given a common past trajectory. Our method provides likelihood estimates of every predicted trajectory, enabling direct integration in streaming optimization. To the best of our knowledge, this is the first work that considers the problem of multiple head motion prediction for 360° video streaming. We first quantify this uncertainty from the data. We then introduce our discrete variational multiple sequence (DVMS) learning framework, which builds on deep latent variable models. We design a training procedure to obtain a flexible and lightweight stochastic prediction model compatible with sequence-to-sequence recurrent neural architectures. Experimental results on 3 different datasets show that our method DVMS outperforms competitors adapted from the self-driving domain by up to 37% on prediction horizons up to 5 sec., at lower computational and memory costs. Finally, we design a method to estimate the respective likelihoods of the multiple predicted trajectories, by exploiting the stationarity of the distribution of the prediction error over the latent space. Experimental results on 3 datasets show the quality of these estimates, and how they depend on the video category.