Over-the-shoulder shot detection in art films

@article{Svanera2015OvertheshoulderSD,
  title={Over-the-shoulder shot detection in art films},
  author={Michele Svanera and Sergio Benini and Nicola Adami and Riccardo Leonardi and Andr{\'a}s B{\'a}lint Kov{\'a}cs},
  journal={2015 13th International Workshop on Content-Based Multimedia Indexing (CBMI)},
  year={2015},
  pages={1-6}
}
The ability to characterize a film, in terms of its narrative and style, is becoming a necessity especially for developing personal video recommendation systems to better deliver on-demand Internet streaming media. Among the set of identifiable stylistic features which play an important role in the film's emotional effects, the use of Over-the-shoulder (OtS) shots in movies is able to convey a big dramatic tension on the viewers. In this work we propose a methodology able to automatically… 

Figures and Tables from this paper

Shot scale distribution in art films

This work proposes an automatic framework for estimating the SSD of a movie by using inherent characteristics of shots containing information about camera distance, without the need to recover the 3D structure of the scene.

Who is the director of this movie? Automatic style recognition based on shot features

This is the first study dealing with automatic attribution of movie authorship, and suggests that feature sequential patterns, i.e. how features evolve in time, are at least as important as the related feature distributions.

Cinematographic Shot Classification through Deep Learning

A data-driven methodology to automatically classify cinematographic shots through deep learning techniques is presented and three different scenarios in which the proposed work can be helpful are discussed.

Analyzing Elements of Style in Annotated Film Clips

This paper presents an open database of annotated film clips together with an analysis of elements of filmic style related to how the shots are composed, how the transitions are performed between

Thinking Like a Director

The evaluation suggests that users generally appreciate the idea of FEPs, and that it can effectively help novice and medium experienced users in crafting film sequences with little training.

Movies and meaning: from low-level features to mind reading

This work addresses the challenge of producing decoding models for the reconstruction of perceptual contents by combining fMRI data and deep features in a hybrid model able to predict specific video object classes by exploiting the intermediate representation of the connotative properties of movies.

Enabling Automatic Cinematography with Reinforcement Learning

A novel unified Reinforcement learning-based Text to Animation (RT2A) framework that can apply reinforcement learning to automatic cinematography and shows that the proposed RT2A can effectively imitate the directors' usage of lens language patterns.

Figaro, hair detection and segmentation in the wild

This work proposes a novel multi-class image database for hair detection in the wild, called Figaro, which tackles the problem of hair detection without relying on a-priori information related to head shape and location.

Quantifying Conflicts in Narrative Multimedia by Analyzing Visual Storytelling Techniques

This study aims at measuring conflict degrees of each shot in visual narrative multimedia by analyzing visual storytelling techniques, such as camerawork, by examining correlations of the proposed conflict measurement with conflict degrees annotated by scholars and practitioners in film studies.

Propriétés du discours de la caméra virtuelle

This thesis concerns the discourse properties of cinematographic storytelling around three axes: time, story, and visual presentation. We address the question of how to analyse and gain knowledge

References

SHOWING 1-10 OF 20 REFERENCES

Shot scale distribution in art films

This work proposes an automatic framework for estimating the SSD of a movie by using inherent characteristics of shots containing information about camera distance, without the need to recover the 3D structure of the scene.

Taxonomy of Directing Semantics for Film Shot Classification

  • H. L. WangL. Cheong
  • Computer Science
    IEEE Transactions on Circuits and Systems for Video Technology
  • 2009
A novel Markov random field based motion segmentation algorithm with an integral foreground/background identification capability based on edge occlusion reasoning is developed, sufficiently robust and fast for film domain conditions, and allows for salient and novel motion descriptors capable of mapping to the proposed directing semantics.

Talking Heads: Detecting Humans and Recognizing Their Interactions

A new learnable context aware configuration model for detecting sets of people in TV material that predicts the scale and location of each upper body in the configuration and substantially outperforms a Deformable Part Model for predicting upper body locations in video frames.

Classification of Cinematographic Shots Using Lie Algebra and its Application to Complex Event Recognition

This paper proposes a discriminative representation of a video shot based on its camera motion and demonstrates how the representation can be used for high level multimedia tasks like complex event recognition and extends its applicability to recognize complex events in large scale videos under unconstrained scenarios.

Using context saliency for movie shot classification

Context saliency is introduced to measure visual attention distributed in keyframes for movie shot classification and has three unique features: 1) context saliency significantly improvesMovie shot classification; 2) the method works for all movie genres; and 3) the methods deals with the most common types of video shots in movies.

Robust real-time attention-based head-shoulder detection for video surveillance

This paper introduces a robust and rapid head-shoulder detection method, which is invariant to pose and viewpoint for video surveillance, and obtains highly competitive results in terms of effectiveness and efficiency.

Shot type characterization in 2D and 3D video content

A variety of possible shot type characterizations that can be assigned in a single video frame or still image are studied and ways to propagate these characterizations to a video segment or to an entire shot are discussed.

Shot type identification of movie content

This paper aims at providing a quantitative description of shot types commonly used in movie productions by generating a database and testing the proposed approach on the set of samples providing promising results.

Head-shoulder detection using joint HOG features for people counting and video surveillance in library

A hierarchical classification method is applied using Haar features and HOG features to head-shoulder location detection and a combined feature named Joint HOG is defined based on the symmetry of head- shoulder portion.

Shot Classification Using Domain Specific Features for Movie Management

Experimental results show that the proposed method classifies movie shots from wide range of movie genres with improved accuracy compared to existing work.