discussion / Camera Traps  / 27 September 2018

Data Needed: Mitigating rhino poaching through multimedia data mining

Greetings everyone, I am a PhD student at Beihang University in Beijing, China. My focusing is in mitigating rhino poaching using multimedia data mining techniques. In order for me to proceed, I should have video dataset which has rhino community. Will be using object detection and abnormal detection methods in order to attempt to develop an algorithm to protect the rhinos. My interest will be getting data, particular from an African context where possibly and if not any video dataset can still be welcomed. Please write to me or advise where I can source out this kind of data I am looking at [email protected] Hope you all receive this in order. Regards Sibusiso Bakana




Hi Sibusiso, 

I'm interested to hear more about your project. From what you've written here, it sounds like you're developing a tool/algorith that will recognise rhino in videos (but not images?), is that right? In what context do you envisge this tool being used (and by whom)? 

In terms of other community members working in this sapce, you could check out @nksheridan 's elephant detection project as well, he mgiht be someone to share ideas with. He made a similar call for training datasets here

Given the focus of your project, it may well be eligible for an AI for Earth innovation grant, so if you're looking for funding you might want to check it out. If your project is not yet ready to put together a proposal, I imagine they are going to have subsequent rounds of funding open after this one, so it would one to keep an eye as you progress. 

Steph 

 

Hi Steph,

Yes, my project objective is to detect rhinos and invaders at the sametime through videos not images. The tool is to be used by the game farm owners in order to protect the rhinos, well by invaders I imply the poachers. 

Thank you very much for the information, yes I need funding as a realtime data will be advantageous for the better prediction. As we all aware mining videos is challenging as compared to images as videos are un-structured data.

Keep in touch.

Sibusiso

Greetings Everyone,

Last week I have written to this community with regards to video data for attempting to analyse videos with rhino community, in order to conduct my research for object detection, object behaviour analysis and invader detection in a protected game farm. Seems It might take time to get such data, I wish to state it clear that the idea is to understand object detection, invader detection and behaviour analysis for wildlife animals and I focused on rhinos because thats what is dangered in my country (South Africa).

Nontheless, if now we dont have specific videos for rhinos, I still welcome video data for elephants. If again there are no video data, I can work on the images of both rhinos and elephants. As we know videos are just a set of continous images linked. For me to conduct my PhD research I depend on data and surely as the project progress, maybe we might find one or two videos that can use to justify my solution. So in simple, anyone who has elephant or rhino images can assist, as both of this wild animals are attcked by poachers for various reasons.

As the objective is to mitigate the poaching of the wild animals through multimedia data mining. 

Please assist a fellow researcher...:)

Thanking you in advanced.

 

Regards

Sibusiso

Hi Sibusiso,

Are you looking for huge data sets or would vacation pictures (mostly Uganda) of elephants be of any help to you?

Cheers, Nigel