Cross-channel Communication Networks

Authors: Jianwei Yang, Zhile Ren, Chuang Gan, Hongyuan Zhu, Devi Parikh

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on multiple vision tasks show that our proposed block brings improvements for different CNN architectures, and learns more diverse and complementary representations.
Researcher Affiliation Collaboration 1Georgia Institute of Technology, 2Facebook AI Research, 3 MIT-IBM Watson AI Lab, 4Institute for Infocomm Research, A*Star, Singapore
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper states: 'We refer to [26] for the implementation of Faster R-CNN. https://github.com/jwyang/faster-rcnn.pytorch, 2017.' However, it does not explicitly provide a link to the open-source code for the methodology described in this paper (the C3 block itself), nor does it state that their code will be released.
Open Datasets Yes We conduct experiments on two popular benchmarks: 1) CIFAR-100 [16]... (2) Image Net [21]... We use Faster R-CNN [20] for object detection on the COCO dataset [19], and Deeplab-V2 [2] for semantic segmentation on the Pascal VOC dataset [5].
Dataset Splits Yes Image Net [21], which has 1000 classes and more than 1.28M images for training, and 50K for validation.
Hardware Specification No The paper does not provide specific hardware details (e.g., exact GPU/CPU models, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper does not provide specific ancillary software details, such as library names with version numbers, needed to replicate the experiments.
Experiment Setup Yes Specifically, we use stochastic gradient descent (SGD) with an initial learning rate 0.1, momentum 0.99, and weight decay 1e-4 for both datasets. The learning rate is decayed by 10 after 100 and 140 epochs for CIFAR-100, and 30 and 60 for Image Net. We report the average best accuracy of 5 runs.