Implement a computer Vision Paper in pytorch

Closed Posted 5 years ago Paid on delivery
Closed Paid on delivery

Surveillance videos are able to capture a variety of realistic anomalies. In this paper, we propose to learn anomalies by exploiting both normal and anomalous videos. To avoid annotating the anomalous segments or clips in training videos, which is very time consuming, we propose to learn anomaly through the deep multiple instance ranking framework by leveraging weakly labeled training videos, i.e. the training labels (anomalous or normal) are at videolevel instead of clip-level. In our approach, we consider normal and anomalous videos as bags and video segments as instances in multiple instance learning (MIL), and automatically learn a deep anomaly ranking model that predicts high anomaly scores for anomalous video segments. Furthermore, we introduce sparsity and temporal smoothness constraints in the ranking loss function to better localize anomaly during training. We also introduce a new large-scale first of its kind dataset of 128 hours of videos. It consists of 1900 long and untrimmed real-world surveillance videos, with 13 realistic anomalies such as fighting, road accident, burglary, robbery, etc. as well as normal activities. This dataset can be usedfortwotasks. First,generalanomalydetectionconsideringallanomaliesinonegroupandallnormalactivitiesin another group. Second, for recognizing each of 13 anomalous activities. Our experimental results show that our MIL method for anomaly detection achieves significant improvement on anomaly detection performance as compared to the state-of-the-art approaches. We provide the results of several recent deep learning baselines on anomalous activity recognition. The low recognition performance of these baselines reveals that our dataset is very challenging and opens more opportunities for future work. The dataset is available at: [login to view URL]

Surveillance videos are able to capture a variety of realistic anomalies. In this paper, we propose to learn anomalies by exploiting both normal and anomalous videos. To avoid annotating the anomalous segments or clips in training videos, which is very time consuming, we propose to learn anomaly through the deep multiple instance ranking framework by leveraging weakly labeled training videos, i.e. the training labels (anomalous or normal) are at videolevel instead of clip-level. In our approach, we consider normal and anomalous videos as bags and video segments as instances in multiple instance learning (MIL), and automatically learn a deep anomaly ranking model that predicts high anomaly scores for anomalous video segments. Furthermore, we introduce sparsity and temporal smoothness constraints in the ranking loss function to better localize anomaly during training. We also introduce a new large-scale first of its kind dataset of 128 hours of videos. It consists of 1900 long and untrimmed real-world surveillance videos, with 13 realistic anomalies such as fighting, road accident, burglary, robbery, etc. as well as normal activities. This dataset can be usedfortwotasks. First,generalanomalydetectionconsideringallanomaliesinonegroupandallnormalactivitiesin another group. Second, for recognizing each of 13 anomalous activities. Our experimental results show that our MIL method for anomaly detection achieves significant improvement on anomaly detection performance as compared to the state-of-the-art approaches. We provide the results of several recent deep learning baselines on anomalous activity recognition. The low recognition performance of these baselines reveals that our dataset is very challenging and opens more opportunities for future work. The dataset is available at: http://crcv.ucf.edu/projects/real-world/

Data Mining Data Science Python

Project ID: #18544010

About the project

5 proposals Remote project Active 5 years ago

5 freelancers are bidding on average $265 for this job

shivampanchal

I have a good hands on working with Advanced R and Python and BI tools and technologies, AI, Big Data. I have quite a good knowledge of DL/ML Algorithm , have also developed Dashboards and Web Applications using flask/ More

$250 USD in 3 days
(104 Reviews)
7.4
dinhfreedom

Dear sir. Your project attracted my attention at first glance, because I've extensive experience in Computer VIsion & PyTorch Programming. I'm really confident about your project, and very eager to join your project. More

$200 USD in 3 days
(40 Reviews)
6.0
arsalan012

Hi, I would like to provide you ny service regarding this project. I'm very passionate about computer vision and I know this would be really interesting project for me. Please try me once and you wouldn't feel regret. More

$277 USD in 15 days
(4 Reviews)
2.9
prayalankar

Hi I have quite of experience working with the tensorflow , keras and Pytorch so coding part I can easily do not a problem. To be true i am also working on the similar kind of problem. I have atm surveillance videos More

$155 USD in 3 days
(2 Reviews)
1.6
oscarBeta

Hi, I have just read through the description and the abstract of the thesis on detection abnormal video clips. It's good to say I have do similar research before on video analytics, using deep learning models. Here are More

$444 USD in 14 days
(0 Reviews)
0.0