Claim · model-written summary
FV encoding can weaken the linear dependency between convolutional activations.
Zea maysNot reportedMaterialGeneCompoundOrgan / stageInteractorInteractor roleSecond interactorSecond interactor roleComparatorSettingValueQualifier
Organisms pictured · Representative pictures of the matched species, not figures from the paper.
Verbatim
e0“targeting four Chinese maize cultivars”
e1“FV encoding can weaken the linear dependency between convolutional activations”
Title
Hao Lu; Zhiguo Cao; Yang Xiao; et al. IEEE Transactions on Automation Science and Engineering 2016. doi:10.1109/tase.2016.2616485
Other claims from this paper (4)
- Method / resource A fine-grained maize cultivar identification data set, HUST-FG-MCI, containing 5000 images was constructed.
- Method / resource The proposed method outperforms other state-of-the-art approaches in maize cultivar identification.
- Observation Redundant filters exist in the convolutional layer.
- Method / resource High accuracy can be maintained with relatively low-dimensional convolutional features and one or two Gaussian components in FV.