A video scene detection algorithm is designed to detect a variety of different scenes within a video

Last update: Jan 04, 2022

Overview

Scene-Change-Detection

The detection of scenes change is a simple problem that human beings face, but it gets much harder to handle autonomously a device that generally includes complex calculations and algorithms.

A video scene detection algorithm is designed to detect a variety of different scenes within a video. There is a very simple definition for a scene: It is a series of logically and chronologically related shots taken in a specific order to depict an over-arching concept or story. The identification of video scenes, in many video analysis applications, is a crucial pre-processing step. A dataset for video scene detection known as the Open Video Scene Detection (OVSD) dataset has been provided in order to evaluate algorithms for video scene detection. Videos in the dataset have an open-source nature, which makes them an ideal product to be used by academics, as well as industry researchers alike.

DATASET

Dataset 2012 - In the dataset, there are six video categories, and in each category, there are four to six video sequences
IBM video Scene Change Detection
A dataset for video scene detection known as the Open Video Scene Detection (OVSD) dataset has been provided in order to evaluate algorithms for video scene detection.

MODEL

VGG16 was used for this project.VGG16 is a convolutional neural network model proposed by K. Simonyan and A. Zisserman from the University of Oxford in the paper “Very Deep Convolutional Networks for Large- Scale Image Recognition”. The model achieves 92.7% top-5 test accuracy in ImageNet, which is a dataset of over 14 million images belonging to 1000 classes.

LIBRARIES USED

Numpy for array manipulation
OpenCV (cv2) for Image Augmentation
Keras for building the Neural Network
Matplotlib for plotting visuals

COMPILATION USED

Loss function selected is sparse categorical cross-entropy
Optimizer selected is Adam
Validation metric chosen is accuracy

Training

No of epochs = 5
Batch size = 1

REPORT

https://drive.google.com/file/d/1cwoP5cRJ5D76PvHV_WjCDRoSJIf0h9du/view?usp=sharing

COLLABORATORS

Neel kumar arya and Ashish Vidyarthi

A video scene detection algorithm is designed to detect a variety of different scenes within a video

Related tags

Overview

Scene-Change-Detection

DATASET

MODEL

LIBRARIES USED

COMPILATION USED

Training

REPORT

COLLABORATORS

License

Owner

Confidence Propagation Cluster aims to replace NMS-based methods as a better box fusion framework in 2D/3D Object detection

[CVPR 21] Vectorization and Rasterization: Self-Supervised Learning for Sketch and Handwriting, IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2021.

Attendance Monitoring with Face Recognition using Python

VisionKG: Vision Knowledge Graph

Official repository of the paper "GPR1200: A Benchmark for General-PurposeContent-Based Image Retrieval"

Generic U-Net Tensorflow implementation for image segmentation

OSLO: Open Source framework for Large-scale transformer Optimization

NEATEST: Evolving Neural Networks Through Augmenting Topologies with Evolution Strategy Training

Code for the paper "Learning-Augmented Algorithms for Online Steiner Tree"

Pytorch and Keras Implementations of Hyperspectral Image Classification -- Traditional to Deep Models: A Survey for Future Prospects.

An Efficient Implementation of Analytic Mesh Algorithm for 3D Iso-surface Extraction from Neural Networks

The code for the NeurIPS 2021 paper "A Unified View of cGANs with and without Classifiers".

A very simple tool for situations where optimization with onnx-simplifier would exceed the Protocol Buffers upper file size limit of 2GB, or simply to separate onnx files to any size you want.

Official public repository of paper "Intention Adaptive Graph Neural Network for Category-Aware Session-Based Recommendation"

An updated version of virtual model making

UniFormer - official implementation of UniFormer

PyTorch3D is FAIR's library of reusable components for deep learning with 3D data

This is the official source code for SLATE. We provide the code for the model, the training code, and a dataset loader for the 3D Shapes dataset. This code is implemented in Pytorch.

A lightweight tool to get an AI Infrastructure Stack up in minutes not days.

Speech recognition tool to convert audio to text transcripts, for Linux and Raspberry Pi.