TEINet: Towards an Efficient Architecture for Video Recognition

Authors: Zhaoyang Liu, Donghao Luo, Yabiao Wang, Limin Wang, Ying Tai, Chengjie Wang, Jilin Li, Feiyue Huang, Tong Lu11669-11676

AAAI 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct extensive experiments to verify the effectiveness of TEINet on several benchmarks (e.g., Something-Something V1&V2, Kinetics, UCF101 and HMDB51). Our proposed TEINet can achieve a good recognition accuracy on these datasets but still preserve a high efficiency.
Researcher Affiliation Collaboration 1State Key Lab for Novel Software Technology, Nanjing University, China 2Youtu Lab, Tencent
Pseudocode No The paper describes its method in text and figures, but does not include structured pseudocode or algorithm blocks.
Open Source Code No The paper does not contain an explicit statement about releasing the source code or a link to a code repository for the described methodology.
Open Datasets Yes Something-Something V1&V2. (Goyal et al. 2017) is a large collection of video clips containing daily actions interacting with common objects. ... Kinetics-400. (Kay et al. 2017) is a large-scale dataset in action recognition... UCF101 (Soomro, Zamir, and Shah 2012) and HMDB51 (Kuehne et al. 2011).
Dataset Splits No The paper does not provide explicit details on how the training, validation, and test splits were created for Something-Something or Kinetics, only mentioning 'three splits' for UCF101/HMDB51 without further definition.
Hardware Specification Yes For all of our experiments, we utilize SGD with momentum 0.9 and weight decay of 1e-4 to train our TEINet on Tesla M40 GPUs using a mini batch size of 64. ... by using a single NVIDIA Tesla P100 GPU to measure the latency and throughput.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., Python, PyTorch, TensorFlow versions).
Experiment Setup Yes On the Kinetics dataset, we train our models for 100 epochs in total, starting with a learning rate of 0.01 and reducing to its 1/10 at 50, 75, 90 epochs. For all of our experiments, we utilize SGD with momentum 0.9 and weight decay of 1e-4 to train our TEINet on Tesla M40 GPUs using a mini batch size of 64.