Exploring Semantic Inter-Class Relationships (SIR) for Zero-Shot Action Recognition
Authors: Chuang Gan, Ming Lin, Yi Yang, Yueting Zhuang, Alexander G.Hauptmann
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Extensive experiments on the UCF101 dataset validate the superiority of our method over fully-supervised approaches using few positive exemplars. |
| Researcher Affiliation | Academia | 1 Institute for Interdisciplinary Information Sciences, Tsinghua University, Beijing, China 2 Centre for Quantum Computation and Intelligent Systems, University of Technology Sydney, Sydney, Australia 3 School of Computer Science, Carnegie Mellon University, Pittsburgh, USA 4 College of Computer Science, Zhejiang University, Zhejiang, China |
| Pseudocode | No | No explicit pseudocode or algorithm blocks were found in the paper. |
| Open Source Code | No | The paper does not explicitly state that its source code is available or provide a link to a code repository for the described methodology. |
| Open Datasets | Yes | To illustrate the effectiveness of the proposed approach, we do experiments on the largest action recognition dataset UCF101 (Soomro, Zamir, and Shah 2012). |
| Dataset Splits | Yes | For all the concept training, where least square regression used, we employ 5-fold cross validations for the value of λ. ... One set contains 91 action classes as known classes for training. The other set contains 10 action classes as unknown classes for testing. ... Thus our testing set consists of 1400 videos of those class actions, while the 12000 videos of the remaining 91 classes can be used for training. |
| Hardware Specification | No | The paper does not specify any particular hardware components (e.g., GPU/CPU models, memory amounts) used for running the experiments. |
| Software Dependencies | No | The paper does not provide specific software dependencies with version numbers, such as Python or PyTorch versions. |
| Experiment Setup | Yes | For all the concept training, where least square regression used, we employ 5-fold cross validations for the value of λ. The search ranges of this parameter are λ {0.01, 0.1, 1, 10, 100}. ... For the compared algorithms, we randomly select 1, 2, 3, 4 and 5 videos as as positive data, and 5000 null videos (collected from Youtube, don t belong to any action class in UCF101) from development set as negative data to train a binary classifier. |