Search

This paper presents a bonobo detection and classification pipeline built from the commonly used machine learning methods. Such application is motivated by the need to test bonobos in their enclosure using touch screen devices without human assistance. This work introduces a newly acquired dataset based on bonobo recordings generated semi-automatically. The recordings are weakly labelled and fed to a macaque detector in order to spatially detect the individual present in the video. Handcrafted features coupled with different classification algorithms and deep-learning methods using a ResNet architecture are investigated for bonobo identification. Performance is compared in terms of classification accuracy on the splits of the database using different data separation methods. We demonstrate the importance of data preparation and how a wrong data separation can lead to false good results. Finally, after a meaningful separation of the data, the best classification performance is obtained using a fine-tuned ResNet model and reaches 75% of accuracy., Comment: IntelliSys 2023 paper

As participants of the MediaEval 2022 Sport Task, we propose a two-stream network approach for the classification and detection of table tennis strokes. Each stream is a succession of 3D Convolutional Neural Network (CNN) blocks using attention mechanisms. Each stream processes different 4D inputs. Our method utilizes raw RGB data and pose information computed from MMPose toolbox. The pose information is treated as an image by applying the pose either on a black background or on the original RGB frame it has been computed from. Best performance is obtained by feeding raw RGB data to one stream, Pose + RGB (PRGB) information to the other stream and applying late fusion on the features. The approaches were evaluated on the provided TTStroke-21 data sets. We can report an improvement in stroke classification, reaching 87.3% of accuracy, while the detection does not outperform the baseline but still reaches an IoU of 0.349 and mAP of 0.110., Comment: Working note paper of the sport task of MediaEval 2022 in Bergen, Norway, 12-13 Jan 2023

This paper presents the baseline method proposed for the Sports Video task part of the MediaEval 2022 benchmark. This task proposes two subtasks: stroke classification from trimmed videos, and stroke detection from untrimmed videos. This baseline addresses both subtasks. We propose two types of 3D-CNN architectures to solve the two subtasks. Both 3D-CNNs use Spatio-temporal convolutions and attention mechanisms. The architectures and the training process are tailored to solve the addressed subtask. This baseline method is shared publicly online to help the participants in their investigation and alleviate eventually some aspects of the task such as video processing, training method, evaluation and submission routine. The baseline method reaches 86.4% of accuracy with our v2 model for the classification subtask. For the detection subtask, the baseline reaches a mAP of 0.131 and IoU of 0.515 with our v1 model., Comment: Baseline paper for the sport Task of MediaEval 2022

Sports video analysis is a widespread research topic. Its applications are very diverse, like events detection during a match, video summary, or fine-grained movement analysis of athletes. As part of the MediaEval 2022 benchmarking initiative, this task aims at detecting and classifying subtle movements from sport videos. We focus on recordings of table tennis matches. Conducted since 2019, this task provides a classification challenge from untrimmed videos recorded under natural conditions with known temporal boundaries for each stroke. Since 2021, the task also provides a stroke detection challenge from unannotated, untrimmed videos. This year, the training, validation, and test sets are enhanced to ensure that all strokes are represented in each dataset. The dataset is now similar to the one used in [1, 2]. This research is intended to build tools for coaches and athletes who want to further evaluate their sport performances., Comment: MediaEval 2022 Workshop, Jan 2023, Bergen, Norway. arXiv admin note: substantial text overlap with arXiv:2112.11384

3D convolutional networks is a good means to perform tasks such as video segmentation into coherent spatio-temporal chunks and classification of them with regard to a target taxonomy. In the chapter we are interested in the classification of continuous video takes with repeatable actions, such as strokes of table tennis. Filmed in a free marker less ecological environment, these videos represent a challenge from both segmentation and classification point of view. The 3D convnets are an efficient tool for solving these problems with window-based approaches., Comment: Multi-faceted Deep Learning, 2021

Sports video analysis is a prevalent research topic due to the variety of application areas, ranging from multimedia intelligent devices with user-tailored digests up to analysis of athletes' performance. The Sports Video task is part of the MediaEval 2021 benchmark. This task tackles fine-grained action detection and classification from videos. The focus is on recordings of table tennis games. Running since 2019, the task has offered a classification challenge from untrimmed video recorded in natural conditions with known temporal boundaries for each stroke. This year, the dataset is extended and offers, in addition, a detection challenge from untrimmed videos without annotations. This work aims at creating tools for sports coaches and players in order to analyze sports performance. Movement analysis and player profiling may be built upon such technology to enrich the training experience of athletes and improve their performance., Comment: MediaEval 2021, Dec 2021, Online, Germany

This paper presents the baseline method proposed for the Sports Video task part of the MediaEval 2021 benchmark. This task proposes a stroke detection and a stroke classification subtasks. This baseline addresses both subtasks. The spatio-temporal CNN architecture and the training process of the model are tailored according to the addressed subtask. The method has the purpose of helping the participants to solve the task and is not meant to reach stateof-the-art performance. Still, for the detection task, the baseline is performing better than the other participants, which stresses the difficulty of such a task.

This paper presents a table tennis stroke detection method from videos. The method relies on a two-stream Convolutional Neural Network processing in parallel the RGB Stream and its computed optical flow. The method has been developed as part of the MediaEval 2021 benchmark for the Sport task. Our contribution did not outperform the provided baseline on the test set but has performed the best among the other participants with regard to the mAP metric., Comment: MediaEval 2021, Dec 2021, Online, Germany

This paper proposes a fusion method of modalities extracted from video through a three-stream network with spatio-temporal and temporal convolutions for fine-grained action classification in sport. It is applied to TTStroke-21 dataset which consists of untrimmed videos of table tennis games. The goal is to detect and classify table tennis strokes in the videos, the first step of a bigger scheme aiming at giving feedback to the players for improving their performance. The three modalities are raw RGB data, the computed optical flow and the estimated pose of the player. The network consists of three branches with attention blocks. Features are fused at the latest stage of the network using bilinear layers. Compared to previous approaches, the use of three modalities allows faster convergence and better performances on both tasks: classification of strokes with known temporal boundaries and joint segmentation and classification. The pose is also further investigated in order to offer richer feedback to the athletes., Comment: MMSports '21, October 20, 2021, Virtual Event,, Oct 2021, Chengdu, China

The paper addresses the problem of recognition of actions in video with low inter-class variability such as Table Tennis strokes. Two stream, "twin" convolutional neural networks are used with 3D convolutions both on RGB data and optical flow. Actions are recognized by classification of temporal windows. We introduce 3D attention modules and examine their impact on classification efficiency. In the context of the study of sportsmen performances, a corpus of the particular actions of table tennis strokes is considered. The use of attention blocks in the network speeds up the training step and improves the classification scores up to 5% with our twin model. We visualize the impact on the obtained features and notice correlation between attention and player movements and position. Score comparison of state-of-the-art action classification method and proposed approach with attentional blocks is performed on the corpus. Proposed model with attention blocks outperforms previous model without them and our baseline.

Heat flow in nanomaterials is an important area of study, with both fundamental and technological implications. However, little is known about heat flow in two-dimensional (2D) devices or interconnects with dimensions comparable to the phonon mean free path (mfp). Here, we find that short, quarter-micron graphene samples reach ~35% of the ballistic heat conductance limit up to room temperature, enabled by the relatively large phonon mfp (~100 nm) in substrate-supported graphene. In contrast, patterning similar samples into nanoribbons (GNRs) leads to a diffusive heat flow regime that is controlled by ribbon width and edge disorder. In the edge-controlled regime, the GNR thermal conductivity scales with width approximately as ~W^{1.8+/-0.3}, being about 100 W/m/K in 65-nm-wide GNRs, at room temperature. Manipulation of device dimensions on the scale of the phonon mfp can be used to achieve full control of their heat-carrying properties, approaching fundamentally limited upper or lower bounds., Comment: Nature Communications, DOI: 10.1038/ncomms2755 (2013)

We study the evolution of non-periodic cosmic string loops containing Y-junctions, such as may form during the evolution of a network of (p,q) cosmic superstrings. We set up and solve the Nambu-Goto equations of motion for a loop with junctions, focusing attention on a specific static and planar initial loop configuration. After a given time, the junctions collide and the Nambu-Goto description breaks down. We also study the same loop configuration in a U(1)xU(1) field theory model that allows composite vortices with corresponding Y-junctions. We show that the field theory and Nambu-Goto evolution are remarkably similar until the collision time. However, in the field theory evolution a new phenomenon occurs: the composite vortices can unzip, producing in the process new Y-junctions, whose separation may grow significantly, destabilizing the configuration. In particular, an initial loop with two Y-junctions may evolve to a configuration with six Y-junctions (all distant from each other). Setting up this new configuration as an initial condition for Nambu Goto strings, we solve for its evolution and establish conditions under which it is stable to the decay mode seen in the field theory case. Remarkably, the condition closely matches that seen in the field theory simulations, and is expressed in terms of simple parameters of the Nambu-Goto system. This implies that there is an easy way to understand the instability in terms of which region of parameter space leads to stable or unstable unzippings., Comment: 16 pages, 11 figures, typos corrected

We present a novel approach for computing the surface roughness-limited thermal conductivity of silicon nanowires with diameter D < 100 nm. A frequency-dependent phonon scattering rate is computed from perturbation theory and related to a description of the surface through the root-mean-square roughness height Delta and autocovariance length L. Using a full-phonon dispersion relation, we find a quadratic dependence of thermal conductivity on diameter and roughness as (D/\Delta)^2. Computed results show excellent agreement with experimental data for a wide diameter and temperature range (25-350 K), and successfully predict the extraordinarily low thermal conductivity of 2 W.m^{-1}.K^{-1} at room temperature in rough-etched 50 nm silicon nanowires., Comment: 12 pages, 4 figures

Current research on the contributions of neurobiology to the understanding of language, and more specifically of phonics, is reviewed. After an introductory section, the second section provides a brief historic perspective on neurolinguistic research, beginning in the early nineteenth century. The third section focuses on what is known about the connection between the evolution of the human brain and the development of human language. Section four describes the physical composition and functioning of the brain, and the subsequent section identifies and discusses studies of the language zones. Issues specific to aphasiology, as distinguished from general neurolinguistics, are then discussed, and related technical terminology is defined. Major research in the last 20 years concerning the phonic aspects of neurolinguistics is reviewed, and the final section offers a model of the neurological function of phonics. A substantial bibliography is appended. (MSE)