Learning Spherical Convolution for Fast Features from 360° Imagery

Authors: Yu-Chuan Su, Kristen Grauman

NeurIPS 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We validate our approach compared to several alternative methods in terms of both raw CNN output accuracy as well as applying a state-of-the-art flat" object detector to 360 data. We evaluate SPHCONV on the Pano2Vid [35] and PASCAL VOC [9] datasets, both for raw convolution accuracy as well as impact on an object detection task.
Researcher Affiliation Academia Yu-Chuan Su Kristen Grauman The University of Texas at Austin
Pseudocode No The paper does not include any pseudocode or clearly labeled algorithm blocks.
Open Source Code No The paper references third-party open-source projects like py-faster-rcnn used in their work, but it does not provide an explicit statement or link for the code implementation of their own proposed methodology.
Open Datasets Yes We use two datasets: Pano2Vid for training, and Pano2Vid and PASCAL for testing. Pano2Vid: We sample frames from the 360 videos in the Pano2Vid dataset [35] for both training and testing. PASCAL VOC: Because the target model was originally trained and evaluated on PASCAL VOC 2007, we 360-ify it to evaluate the object detector application.
Dataset Splits No The paper states the number of frames for training (1,056) and testing (168) from the Pano2Vid dataset, but it does not specify a separate validation dataset split.
Hardware Specification No The paper does not provide specific details regarding the hardware used for running the experiments, such as GPU models, CPU types, or cloud computing specifications.
Software Dependencies No The paper mentions software like VGG, Faster-RCNN, and Caffe, but it does not specify version numbers for these or any other software dependencies needed to replicate the experiments.
Experiment Setup Yes We set the resolution of Ie to 640 320. For the projection operator P, we map α=65.5 to W=640 pixels following SUN360 [38]... The kernel size upper bound Uk=7 7 following the max kernel size in VGG. We insert batch normalization for conv4_1 to conv5_3.