Artificial Intelligence is used in computer science and does tasks that usually require human intelligence. This can include various fields such as decision-making. These machines are becoming more advanced with time and one of its latest accomplishments is being able to tell how many people there are in large crowds much faster and more accurate than ever before.
Prior, experts would have to estimate the amount of people in incredibly large crowds. However machines can now do the work for us. Machines were able to do this before but it was not very accurate and took just the same amount of time as if a human were to do it. But now, this new system has proven to be 15% more accurate over any other crowd estimator. Furthermore, it is extremely fast and takes almost no time at all to calculate, especially compared to when humans try to. Having this machine can be very useful, especially in situations deemed dangerous once a certain capacity is reached. An example can be a mosh pit at a concert. Moreover, in history, years from now having this type of accurate data can be very useful.
This seems like an interesting use of AI that I haven’t heard of before. To some this may not seem like as useful as AI being used to drive cars or perform factory tasks, but I see this as a potentially very useful tool. It could be used for places that have max capacity issues where too many people are in one spot and can turn into a dangerous situation. This could also play a big role in todays society as mass shootings seem to be trending in the news. With a few advancements in this technology, say being able to estimate and spot people in crowds, it could be a useful tool for police forces and investigators. It could help either identify a suspect or at least help investigators get pieces of the puzzle to mass shooting situations or of the like. It could be helpful finding how many people were in a specific area of an attack. All of this could be potentially very useful information for analyzing such situations and finding out what happened much quicker than before. This would certainly be useful as looking at a picture and trying to count the amount of people, let alone trying to find specific people would be very time consuming and difficult. It seems as though this advancement in this AI technology has a lot to offer, especially as it advances and upgrades.
This is really interesting! How do you think it is able to do this? Does it use sensors at the entrances / exits to track how many people come and go? Or does it count each person individually from a birds-eye-view?
I just looked it up and found some cool tools we used to use before this to get a rough estimate. I’ll leave the link below. This tool essentially brought up a map where you can then outline the space the ‘event’ is / has taken place. Estimate the crowd density, listing how many people you think occupy each square meter of space and then search! The website will then estimate how many people are in that crowd based off of the location and density. The tool you are describing however seems much more technologically advanced and would be able to report more precise values as opposed to solely a rough estimate.
Online tool:
https://www.mapchecking.com/
References:
https://www.poynter.org/fact-checking/2017/this-online-tool-makes-checking-crowd-sizes-easier/
The artificial intelligence opinion that you mention is pretty interesting. Artificial intelligence did better than human-being in the lots of field like calculate, medical, decision-making and so on. there was news that AlphaGo defeated Ke Jie who is the world’s number one Go player. As Sam Byford (2017) mentioned “Ke played ‘perfectly’ for about the first 50 moves, according to AlphaGo’s analysis, and evolved his strategy to engage in a series of complex battles across the board, but AlphaGo handled the multi-angled assault well to simplify the engrossing game and eventually forced Ke to resign.”
As we can see here, when AL faces a complex situation, they will make things more simple, however, when human face a situation, typically, lots of human will make thing more strategy, but when human want to solve it, finally, they cannot how to solve it. Therefore, Is it human different strategies with AL when they both solve situations?
https://www.theverge.com/2017/5/25/15689462/alphago-ke-jie-game-2-result-google-deepmind-china
Your post on Artificial Intelligence has inspired me to think about how Artificial Intelligence would have a positive impact on using it as a crowd estimator. One of the most interesting applications of AI in using crowd estimator to detect gas leaks, and subway wait times. As mentioned on the website (https://www.scmp.com/tech/start-ups/article/3019425/hong-kong-start-behind-protest-crowd-counting-using-ai-estimate), AI can be highly beneficial in times of protest such as in Hong Kong, when the other citizens of the country wanted to make use of the available subway stations and continue their daily lifestyle without being affected by the public protests. Hence, this might come in handy there and it did. Such situations dictate the better use of AI in crowd estimation.
However, there is always a potential threat of that system of such a public scale being hacked by hackers and causing more traffic than a solution.
Source: https://www.scmp.com/tech/start-ups/article/3019425/hong-kong-start-behind-protest-crowd-counting-using-ai-estimate
Super interesting application of artificial intelligence! I always love seeing new and interesting ways people are using artificial intelligence to make tasks and everyday lives easier. This post reminded me of when we learned about China’s facial recognition software it was using on its citizens. I could see both of these technologies being used together. The facial recognition software could be used to help monitor large crowds and keep them safe. The Chinese recognition software we learned about was able to bring up information on the people in real time as they cross the street.
In recent times, many large scale events have been under high security due to mass acts of terror. Bombings, shootings and other acts of terror have become a big concern whenever there are events in which large crowds congregate.The facial recognition software would be able to detect possibly threats in the crowd and prevent possibly harmful events.
Source:https://www.digitaltrends.com/cool-tech/facial-recognition-china-50000/