A Laplace-inspired Distribution on SO(3) for Probabilistic Rotation Estimation

Authors: Yingda Yin, Yang Wang, He Wang, Baoquan Chen

ICLR 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Our extensive experiments show that our proposed distribution achieves state-of-the-art performance for rotation regression tasks over both probabilistic and non-probabilistic baselines.
Researcher Affiliation Academia Yingda Yin Yang Wang He Wang Baoquan Chen Peking University
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No Our project page is at pkuepic.github.io/Rotation Laplace.
Open Datasets Yes Model Net10-SO3 (Liao et al., 2019) is a commonly used synthetic dataset for single image rotation estimation containing 10 object classes.
Dataset Splits No For Pascal3D+ dataset, it is mentioned that images are split into Image Net train, Image Net val, Pascal VOC train, and Pascal VOC val sets. For Pascal3D+ dataset, we follow Murphy et al. (2021) to use (the more challenging) Pascal VOC val as test set. However, the paper does not provide explicit percentages or counts for the train/validation/test splits used in their experiments.
Hardware Specification Yes The experiment is done on Model Net10-SO3 toilet dataset on a single 3090 GPU.
Software Dependencies No The paper mentions using 'pretrained Res Net-101' and 'SGD optimizer' but does not specify software versions for libraries, frameworks, or programming languages used.
Experiment Setup Yes The batch size is set as 32. We use the SGD optimizer and start with the learning rate of 0.01. For Model Net10-SO3 dataset, we train 50 epochs with learning rate decaying by a factor of 10 at epochs 30, 40, and 45. For Pascal3D+ dataset, we train 120 epochs with the same learning rate decay at epochs 30, 60 and 90.