Poster-Based Multiple Movie Genre Classification Using Inter-Channel Features

Abstract

As the scale of the film industry grows, the demand for well-established movie databases is also growing. The genre of a movie supplies information on its overall content and has multiple values. Therefore, it should be well classified utilizing the characteristics of movies, without omissions in the database. In this study, we extract the optimal information and characteristics from movie posters to aid the classification of movies into genres and propose the use of a Gram layer in a convolutional neural network (CNN). The Gram layer first extracts style features by applying the Gram matrix to produce a feature map of a poster image. Using this as a style weight, the existing feature map is merged with style information to perform the genre classification task. The proposed Gram layer performed multi-genre classification tasks with higher efficiency than a residual neural network (ResNet), which is the current CNN model used for such tasks. We compared the activation map with the Squeeze-and-Excitation network, which gives weight to the image, and we confirmed that the introduction of the Gram layer actually focuses on the style of the movie poster. To classify the movie genres, we reconstructed the poster dataset into 12 multi-genres that emphasized the characteristics of each poster.

Genre Classification of Telugu and English Movie Based on the Hierarchical Attention Neural Network

“…The results present in the Table 4 indicate that the proposed HANN method has achieve better performance in terms of accuracy and precision when compared with gram layer in CNN [16]. The HANN method has achieved less accuracy (only 73.15%) compared to DCNN [17], this is because the HANN method uses the text as input data and the imbalance and skewed data are presented in the collected movie data.…”

“…Wi, Jang and Kim [16] developed a gram layer in CNN to extract the optimal information from the movie posters for final classification. A feature map was created by applying the gram matrix with style features.…”

“…In this study, a fully-connected Gated Recurrent Unit (GRU) network [16] known as GRU2 and a bidirectional GRU is used. The temporal information of forward or backward at a time is presented in the bidirectional GRU as BiGRU.…”

“…The performance of the HANN method is compared with the gram layer in CNN [16] and DCNN [17] in terms of accuracy and precision on text and video sequence data.…”

“…The precision of HANN method has achieved 93% compared to the both existing techniques, because of the inclusion of the attention layer in the proposed method. The existing Gram layer [16] achieved poor recall (i.e. 47.22%), because recall is only calculated for label values and didn't considered the sample values for both input and output data.…”