Memory Attention Networks for Skeleton-based Action Recognition

Authors: Chunyu Xie, Ce Li, Baochang Zhang, Chen Chen, Jungong Han, Jianzhuang Liu

IJCAI 2018 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The proposed MANs are evaluated on four public skeleton action datasets: NTU RGB+D [Shahroudy et al., 2016], HDM05 [M uller et al., 2005], SYSU-3D [Hu et al., 2015] and UT-Kinect [Xia et al., 2012].
Researcher Affiliation Collaboration 1 School of Automation Science and Electrical Engineering, Beihang University, Beijing, China 2 China University of Mining & Technology, Beijing, China 3 Department of Electrical and Computer Engineering, University of North Carolina at Charlotte 4 School of Computing & Communications, Lancaster University, LA1 4YW, UK 5 Noah s Ark Lab, Huawei Technologies Co. Ltd., China
Pseudocode No The paper does not include any structured pseudocode or algorithm blocks.
Open Source Code Yes The code will be made publicly available at https://github.com/memory-attention-networks.
Open Datasets Yes The proposed MANs are evaluated on four public skeleton action datasets: NTU RGB+D [Shahroudy et al., 2016], HDM05 [M uller et al., 2005], SYSU-3D [Hu et al., 2015] and UT-Kinect [Xia et al., 2012].
Dataset Splits Yes For all the datasets, the matrices (e X, e Y and e Z) are generated with all the frames of a skeleton sequence. For a fair comparison, the performance of MANs on each dataset is compared with existing methods using the same evaluation protocol. The mini-batches of samples on NTU RGB+D, HDM05, SYSU-3D, and UT-Kinect are constructed by randomly sampling 40, 20, 8, and 8 samples from the training sets, respectively.
Hardware Specification Yes All experiments are performed based on Keras2 with Tensorflow backend using two NVIDIA Titan X Pascal GPUs.
Software Dependencies No The paper mentions using "Keras2 with Tensorflow backend", but does not specify the version numbers for Keras or Tensorflow.
Experiment Setup Yes For the large scale, the number of hidden units of Bi GRU in TARM is set to 2 128 (K = 128), where 2 indicates bidirectional GRU, 128 is the number of neurons. MANs are trained using the stochastic gradient descent algorithm, and the learning rate, decay, and momentum, are respectively set to 0.1, 0, and 0.9. The mini-batches of samples on NTU RGB+D, HDM05, SYSU-3D, and UT-Kinect are constructed by randomly sampling 40, 20, 8, and 8 samples from the training sets, respectively. The training stops after 100 epochs except for NTU RGB+D after 50 epochs.