Combo-Action: Training Agent For FPS Game with Auxiliary Tasks

Authors: Shiyu Huang, Hang Su, Jun Zhu, Ting Chen954-961

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

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments show that our method is efficient in training process and outperforms previous state-of-the-art approaches by a large margin. Ablation study experiments also indicate that our method can boost the performance of the FPS agent in a reasonable way.
Researcher Affiliation Academia Shiyu Huang, Hang Su, Jun Zhu, Ting Chen Dept. of Comp. Sci. & Tech., BNRist Center, State Key Lab for Intell. Tech. & Sys., Institute for AI, THBI Lab, Tsinghua University, Beijing, 100084, China hsy17@mails.tsinghua.edu.cn; {suhangss, dcszj, tingchen}@tsinghua.edu.cn
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an unambiguous statement that the authors are releasing the code for the work described in this paper, nor does it provide a direct link to a source-code repository.
Open Datasets No The paper states 'We collect a set of labeled images for training and testing from the Vi ZDoom environment.' and 'In this paper, we collected 10,000 images for training and use extra 2000 images as validation dataset.' but does not provide access information (link, DOI, repository, or citation) for this dataset.
Dataset Splits Yes We then split our dataset into 3 partitions: Train: Validate :Test, with ratios 70%:20%:10%.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions algorithms and models but does not provide specific ancillary software details like library names with version numbers (e.g., Python, TensorFlow, PyTorch versions).
Experiment Setup Yes Our learning rate schedule is as follows: For the first 195, 000 steps we start at a high learning rate 10−3. Then we continue training with 10−4 for 3, 000 steps, and finally 10−5 for 2, 000 steps. Dropout (Srivastava et al. 2014) with ratio 0.5 is used during training.