General Ideas Generated Using The Technologies Learnt During The IIEC DOT 2 Day Workshop On AI
In this workshop, we have learnt a lot of things from basics of linux and python to Computer Vision and Mask RCNN ML algorithm. These aspects can be integrated together to help fight the current pandemic and also aid those people who are working hard for us everyday ( doctors and police for example ). Some of the best ideas were already put forth by Vimal sir. These aspects can also be used in general for other purposes and not just for fighting COVID-19.
- By using the face recognition algorithm, we can find the criminals who are hidden. If we know how the criminal looks, we can make a picture of him/her and give that picture as input to the algorithm. We can then attach cameras to drones which travels around and upon spotting the criminal sends us a signal indicating the same. We can also find out missing people using the same concept.
- *IMPORTANT :- In places where a human cannot go due to unfitting survival conditions but he/she has a need to explore the place on the basis of research to find an object then, Mask RCNN can be used to mask the objects in that place and find out if the object that the person is looking for is present or not. I can state an example for this kind of situation. When the astrobiologists have to explore the extreme conditioned places on earth to find out what kind of living organisms or extremophiles live there, instead of risking their lives, they can send drones which can sustain very high or very low temperatures embedded with Mask RCNN algorithm to capture images and mask the shape of organisms. Scientists can then analyse and study the body structures of those organisms. This same method can be used to find extraterrestrial life on planets like Mars when the rovers capture images from a long distances, the objects in the images can be masked.
- *IMPORTANT :- Mask RCNN or just the image recognition can be used to find missing ships on the surface of the sea or missing submarines in the underwater depths.
- *IMPORTANT :- Another important application can be the use of Mask RCNN to locate the sunspots or any other star spots to study to study the interaction of the star's surface plasma with the magnetic field it generates.
- *IMPORTANT :- From the far distance images captured by space telescopes like kepler or TESS, scientists can differentiate between starts, planets and other stuff present in the cosmos using CV and Mask RCNN
Firstly, using a single picture and training the model to identify the person is not that effective. You may get both False Positives as well as True Negatives of which both are harmful as the accused can be left and an innocent ca be declared guilty. Although you can use that algorithm effectively, such technology of finding criminals and other places where people can't reach has already been developed and making good progress. I feel like we can't identify minute objects using the RCNN algorithm. And with the shape and structure of organisms, we cannot study an organism. It is a good article to refer to. All the best!