Optimizing multimodal scene recognition through relevant feature selection approach for scene classification.

Scene classification plays a vital role in various computer vision applications, but building deep learning models from scratch is a very time-intensive process. Transfer learning is an excellent classification method using the predefined model. In our proposed work, we introduce a novel method of multimodal feature extraction and a feature selection technique to improve the efficiency of transfer learning in scene classification.

Genetic characterization of fall armyworm infesting South Africa and India indicate recent introduction from a common source population.

The invasion of the Western Hemisphere native fall armyworm (Spodoptera frugiperda; J. E. Smith) (Lepidoptera: Noctuidae) into the Eastern Hemisphere has been notable for the rapidity and geographical breadth of new detections.

Marker-assisted improvement of the elite restorer line of rice, RPHR-1005 for resistance against bacterial blight and blast diseases.

This study was carried out to improve the RPHR-1005, a stable restorer line of the popular medium slender grain type rice hybrid, DRRH-3 for bacterial blight (BB) and blast resistance through marker-assisted backcross breeding (MABB). Two major BB resistance genes, Xa21 and Xa33 and a major blast resistance gene, Pi2 were transferred to RPHR-1005 as two individual crosses. Foreground selection for Xa21, Xa33, Pi2, Rf3 and Rf4 was done by using gene-specific functional markers, while 59 simple sequence repeat (SSR) markers polymorphic between the donors and recipient parents were used to select the best plant possessing target resistance genes at each backcross generation.