Perceptual Quality Assessment of Omnidirectional Images

Authors: Yuming Fang, Liping Huang, Jiebin Yan, Xuelin Liu, Yang Liu580-588

AAAI 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experimental results on our database verify that the proposed model achieves the competing performance compared with the state-of-the-art methods.
Researcher Affiliation Collaboration Yuming Fang1, Liping Huang1, Jiebin Yan1*, Xuelin Liu1, Yang Liu2 1 Jiangxi University of Finance and Economics, Nanchang, China 2 SANY Heavy Industry Co., Ltd., China
Pseudocode No No pseudocode or algorithm blocks labeled as such were found in the paper. The paper presents a framework diagram and mathematical equations but not structured code.
Open Source Code No The paper does not provide any concrete access to source code, such as a repository link, nor does it explicitly state that the code will be released or is available in supplementary materials.
Open Datasets No The paper describes the construction of 'our database' but does not provide any access information (link, DOI, citation) for this specific dataset to be publicly available.
Dataset Splits Yes We use 80% omnidirectional images of the proposed database for training and the rest 20% for testing. We take 10 times of random train-test splitting operation and report the median performance to reduce any bias.
Hardware Specification Yes The proposed model is implemented with Py Torch on an NVIDIA Ge Force GTX 1080 Ti machine.
Software Dependencies No The paper mentions 'Py Torch' but does not specify its version number, nor does it list any other software dependencies with their versions.
Experiment Setup Yes We adopt Adam (Kingma and Ba 2017) optimizer with weight decay 5 10 4 and set mini-batch size to 4 for training 100 epochs, where the learning takes roughly a half-day. The initial learning rate is set to 10 4 and reduced by a decay factor 0.1 for 50 epochs.