In India, there has been a notable struggle between the population growth and the wildlife. Injuries, fatalities, destruction of human habitation, crop devastation, and human property damage are only a few of the serious effects. Temporary measures to safeguard the habitat, such as electric fences, trenches, manual surveillance, guard dogs, etc., are used, but they are not cost-effective and have been shown to be harmful for both humans and wildlife. Some sort of mitigation strategy is needed to address this problem in a way that ensures the safety of both wild animals and people. Inspite of many results to monitor animal safety, AI provides additional benefits. An obstacle which revolve around this problem may undoubtedly be pushed forward by using IoT alone. Development of new methodology to identify the incursion of forest creatures after which they need to be sent again to their place safely can be obtained using the introduced idea. Also, this idea helps in human safety against animal attack on them. In the proposed work, we combine YOLOv3 weights and machine learning approaches to address the issue at hand by reducing the time it takes to recognize many objects while maintaining the highest level of time complexity. The pre-defined neural network algorithm referred to as the YOLO framework will be used to process the captured image. When an object is recognized, the processor sends an email updating the presence of animals, and a buzzer turns on when Arduino is activated, simultaneously turning on the speaker with crackers sounds to repel animals. Through the IOT module, it sends the taken photographs to the mail of the authorized person.