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

links and status of cool gradio demos

Notebooks, slides and dataset of the CorrelAid Machine Learning Winter School

NasirKhusraw - The TSP solved using genetic algorithm and show TSP path overlaid on a map of the Iran provinces & their capitals.

Control-Raspberry-Pi-Robot-using-Hand-Gestures - A 4WD Robot car based on Raspberry Pi that controlled by hand gestures(using openCV and mediapipe)

A Python package for generating concise, high-quality summaries of a probability distribution

ULMFiT for Genomic Sequence Data

An implementation for the ICCV 2021 paper Deep Permutation Equivariant Structure from Motion.

Tool for working with Y-chromosome data from YFull and FTDNA

Source Code for AAAI 2022 paper "Graph Convolutional Networks with Dual Message Passing for Subgraph Isomorphism Counting and Matching"

FastFCN: Rethinking Dilated Convolution in the Backbone for Semantic Segmentation.

Object detection (YOLO) with pytorch, OpenCV and python

GenshinMapAutoMarkTools - Tools To add/delete/refresh resources mark in Genshin Impact Map

Reimplementation of Dynamic Multi-scale filters for Semantic Segmentation.

Official Pytorch implementation of paper "Reverse Engineering of Generative Models: Inferring Model Hyperparameters from Generated Images"

PyTorch Implementation for Fracture Detection in Wrist Bone X-ray Images

A real-time speech emotion recognition application using Scikit-learn and gradio

Implementation of Self-supervised Graph-level Representation Learning with Local and Global Structure (ICML 2021).

Neural Re-rendering for Full-frame Video Stabilization

Variational Attention: Propagating Domain-Specific Knowledge for Multi-Domain Learning in Crowd Counting (ICCV, 2021)

Repository for the NeurIPS 2021 paper: "Exploiting Domain-Specific Features to Enhance Domain Generalization".