Study of Spatio-Temporal Modeling in Video Quality Assessment

Fang, Yuming; Li, Zhaoqian; Yan, Jiebin; Sui, Xiangjie; Liu, Hantao

doi:10.1109/tip.2023.3272480

Cited by 7 publications

(1 citation statement)

References 66 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…With a substantially lower computational cost, the algorithm removes temporal redundancy by selecting a set of representative frames and achieves promising performance. In [110], different frame sampling strategies were designed. The findings of this study show that sparsely sampled video frames can obtain a competitive performance against using all video frames for quality estimation.…”

Section: No-reference Quality Assessmentmentioning

confidence: 99%

Efficient Computational Cost Saving in Video Processing for QoE Estimation

Llorente,

Pérez,

Rodrigo

et al. 2024

IEEE Access

View full text Add to dashboard Cite

No-Reference video quality assessment has become a trending and challenging hot topic in estimating perceived quality in audiovisual content. In this paper, we present a proposal to considerably reduce the computational cost of video processing without losing accuracy in QoE estimation. Tests have been performed using the Video-MOS SaaS solution, a hybrid NR-VQA solution based on perceptible video distortions and a machine learning approach. After exploring the spatial and temporal redundancy present in a video sequence, the final approach combines video metric feature extraction in both high and low video resolution, together with a specific frame selection based on a uniform temporal sampling and frame type at the video coding level. An extensive validation with more than 144 hours of audiovisual content from six of the most important HD channels of DTT in Spain demonstrates the validity of the approach, ensuring real-time application on the test device, with computational cost savings of 94.96% and an obtained MOS error of 0.1144, in more than 174000 3-second measurements.INDEX TERMS Computational cost, feature extraction, I frames, machine learning, Mean Opinion Score (MOS), no-reference, video quality assessment, perceived quality, Quality of Experience (QoE), video processing.

show abstract

Section: No-reference Quality Assessmentmentioning

confidence: 99%

Efficient Computational Cost Saving in Video Processing for QoE Estimation

Llorente,

Pérez,

Rodrigo

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

Revisiting the robustness of spatio-temporal modeling in video quality assessment

Yan,

Wu,

Jiang

et al. 2024

Displays

View full text Add to dashboard Cite

The Short Video Popularity Prediction Using Internet of Things and Deep Learning

He,

2024

IEEE Access

View full text Add to dashboard Cite

In order to furnish valuable insights and solutions applicable to content creators, social media platforms, academic research, and general users, this investigation integrates the Internet of Things (IoT) with deep learning regression models to examine methodologies for predicting the popularity of short videos. Within the context of cross-cultural communication, a proposed Content Popularity Rank Prediction based on the Convolutional Neural Network (CPRP-CNN) model relies exclusively on the personal attributes of the publisher and the textual characteristics of short videos to anticipate the viewership levels of short videos promptly following their release. Through simulated experiments, the model's performance is assessed, revealing that the utilization of the Rectified Linear Unit (Relu) activation function in the CPRP-CNN model enhances accuracy by 42.2% when contrasted with the use of the sigmoid function. This enhancement is coupled with a 37.8% reduction in cross-entropy loss. Furthermore, the proposed CPRP-CNN model attains a cross-entropy of 0.692 and an accuracy of 74.7%, exhibiting superior Mean Squared Error (MSE) and Mean Absolute Error (MAE) values of 2.728 and 1.751, respectively, when compared to alternative prediction models. These outcomes signify that the amalgamation of deep learning models with fused features within the IoT context significantly ameliorates the predictive efficacy of short video popularity. The research findings contribute to the enhancement of personalized and precise short video content recommendations.

show abstract

Study of Spatio-Temporal Modeling in Video Quality Assessment

Cited by 7 publications

References 66 publications

Efficient Computational Cost Saving in Video Processing for QoE Estimation

Efficient Computational Cost Saving in Video Processing for QoE Estimation

Revisiting the robustness of spatio-temporal modeling in video quality assessment

The Short Video Popularity Prediction Using Internet of Things and Deep Learning

Contact Info

Product

Resources

About