Uncertainty-aware No-Reference Point Cloud Quality Assessment

2024-01-17

arXiv.org

AI Summary

Existing point cloud quality assessment (PCQA) methods formulate mean opinion score (MOS) prediction as deterministic regression, overlooking the inherent stochasticity of subjective ratings. This work proposes the first uncertainty-aware no-reference PCQA framework tailored for point cloud compression and enhancement, explicitly modeling the distributional characteristics of multi-subject judgments. Our approach introduces: (1) a novel probabilistic no-reference PCQA paradigm; (2) a conditional variational autoencoder (CVAE)-based architecture that jointly models prior, posterior, and quality score generation to emulate the multi-subject rating process; and (3) a diversity-aware multi-score ensemble strategy yielding robust MOS estimates. Extensive experiments demonstrate significant improvements over state-of-the-art methods across multiple benchmarks, with strong cross-dataset generalization capability.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Uncertainty Representations
• Knowledge Representation and Reasoning: Qualitative Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

The evolution of compression and enhancement algorithms necessitates an accurate quality assessment for point clouds. Previous works consistently regard point cloud quality assessment (PCQA) as a MOS regression problem and devise a deterministic mapping, ignoring the stochasticity in generating MOS from subjective tests. Besides, the viewpoint switching of 3D point clouds in subjective tests reinforces the judging stochasticity of different subjects compared with traditional images. This work presents the first probabilistic architecture for no-reference PCQA, motivated by the labeling process of existing datasets. The proposed method can model the quality judging stochasticity of subjects through a tailored conditional variational autoencoder (CVAE) and produces multiple intermediate quality ratings. These intermediate ratings simulate the judgments from different subjects and are then integrated into an accurate quality prediction, mimicking the generation process of a ground truth MOS. Specifically, our method incorporates a Prior Module, a Posterior Module, and a Quality Rating Generator, where the former two modules are introduced to model the judging stochasticity in subjective tests, while the latter is developed to generate diverse quality ratings. Extensive experiments indicate that our approach outperforms previous cutting-edge methods by a large margin and exhibits gratifying cross-dataset robustness.

Problem

• Addresses stochasticity in point cloud quality assessment
• Proposes probabilistic model for no-reference quality evaluation
• Simulates diverse subjective judgments for accurate MOS prediction

Innovation

• Probabilistic architecture for no-reference PCQA
• Conditional variational autoencoder models judging stochasticity
• Generates diverse ratings simulating subjective judgments