Movie genre classification: A multi-label approach based on convolutions through time

Wehrmann, Jônatas; Barros, Rodrigo C.

doi:10.1016/j.asoc.2017.08.029

Cited by 73 publications

(42 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similarly, Wehrmann and Barros [6] attempted genre classification by applying a convolutional network to trailers. In addition, their study used residual connections and built a convolution-through-time network for multi-label movie genre classification.…”

Section: B Non-poster-based Movie Genre Classificationmentioning

confidence: 99%

“…To overcome these problems and perform genre classification efficiently, many previous studies utilized machine learning and deep learning to attempt automatic genre classification, based on various data, such as movie posters [1], [2], plots [3], [4] and trailers [5], [6], which were used separately or in combination. However, the use of movie plots is limited by the fact that they express only the introductory portion of the main plot and not the entire content.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

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

Jang

Kim

2020

IEEE Access

View full text Add to dashboard Cite

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.

show abstract

Section: B Non-poster-based Movie Genre Classificationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

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

Jang

Kim

2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…In addition to the use of low-level descriptors extracted from video, other authors proposed to classify or characterize films using the video content (mainly video frames extracted from trailers) and deep learning techniques. This is the case of [11] and [12] where authors classify films using their trailers, with the application of Convolutional Neural Networks (CNNs), comparing the results with other deep learning techniques such as LSTM (Long short-term memory). Authors in [13] perform the classification with films poster images using CNNs, whereas authors in [14] are using previews.…”

Section: Related Workmentioning

confidence: 99%

On the influence of low-level visual features in film classification

et al. 2019

View full text Add to dashboard Cite

Background In this paper we present a model of parameters to aesthetically characterize films using a multi-disciplinary approach: by combining film theory, visual low-level video descriptors (modeled in order to supply aesthetic information) and classification techniques using machine and deep learning. Methods Four different tests have been developed, each for a different application, proving the model's usefulness. These applications are: aesthetic style clustering, prediction of production year, genre detection and influence on film popularity. Results The results are compared against high-level information to determine the accuracy of the model to classify films without knowing such information previously. The main difference with other film characterization approaches is that we are able to isolate the influence of high-level descriptors to really understand the relevance of low-level features and, accordingly propose a useful set of low-level visual descriptors for that purpose. This model has been tested with a representative number of films to prove that it can be used for different applications.

show abstract

“…One of the drawbacks in their model is a large number of parameters involved along with VGG(about 13 crores), thereby leading to substantial computation time as compared to our model. This paper [10] works on the concept of multiple label system, which enable the model to classify the trailer into more than 1 genre which is realistic in comparison to the single labeled which imparts single genre to all scenes. They used max-pooling and convolution to extract features.…”

Section: Related Workmentioning

confidence: 99%

Multimodal KDK Classifier for Automatic Classification of Movie Trailers

Shambharkar¹,

Doja²,

Chandel³

et al. 2019

IJRTE

View full text Add to dashboard Cite

Movie trailer classification is a field of automation of analyzing the movie trailers and classify them into one of the various genres. In this paper, we proposed a classifier to identify the genre of a movie trailer by analyzing it's audio and visual features simultaneously. Our Approach decomposes a trailer video into frames and audio file and then analyze them based on certain specific features to categorize them into four genres. Our aim was to minimize the number of parameters involved in analyzing the trailer since other papers use many arguments which are impractical. The proposed classifier was trained on 4 audio, 2 broad visual features extracted from over 900 movie trailers distributed across 4 different genres, namely Drama, Horror, Romance, and Action. The Classifier Model has been trained using Neural Networks and Convolutional Neural Networks. Our Classifier Model can be used in Recommendation Systems and various websites like IMDB for automation of the genre classification process. As the common humanly approach is to generalize the results obtained from many inputs, the same way we use multiple models to obtain different outputs from multiple ANN models and then combine all the obtained results to get a final output. Also a Dataset containing 1000 movie trailers was introduced in this paper with trailers spanning to almost all Hollywood movies from 2010-2019 .After training and conducting Experiments on around 1000 movie trailers, the classifier model showed a maximum accuracy of 81 percent in determining top 1 genre and 91 percent in determining top 2 genres of a movie trailers in the test set

show abstract

Movie genre classification: A multi-label approach based on convolutions through time

Cited by 73 publications

References 9 publications

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

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

On the influence of low-level visual features in film classification

Multimodal KDK Classifier for Automatic Classification of Movie Trailers

Contact Info

Product

Resources

About