The objective of this research paper on research productivity in the Indian Automotive Industry is to present the transformation of the research community of the Indian automotive Industry in a progressively responsible way, with reference to SIAT conferences held so far. It reveals the dynamic nature of automotive research in the country. The exhibit's research outcome is dominated by specific domains and specific organizations in a dynamic way. It is also elaborating the countries and scientists preferred to collaborate with and their impact on the Indian automotive Industry.
The world is becoming automated day by day in each and every aspect. The technological approaches to solve every human problem are evolving at a great interest in this rapid advancing environment. Automated houses, cars(for instance, Tesla self-driving cars), machinery, industries can be considered as the progressive outcomes which made the human life comfort, safe and easy. Although, there exist numerous situations that are damaging and the lives of people miserably . One of those circumstances is “Road Accidents”. There may be many reasons for this problem where “Driver fatigue” is one of those. “Each year, around 1.5 lakh people die in road mishaps in India.” –The Times of India[17]. To mitigate these accidents, we develop a model based on Convolutional Neural Networks (CNNs), Deep Learning which can detect the sleepiness or fatigue and notify the driver(or the person who is driving). In this paper, we developed a model using CNN classifier that identifies whether a driver is sleepy or not. The model here also provides an alerting alarm when the driver is sleeping (whenever he/she closes eyes). The proposed model is also evaluated using large amount of data to increase its accuracy and correctness while detecting the person’s drowsiness.
Making insightful systems to prevent minor accidents can be particularly incredible in limiting the difficulty and loss of life. Street crashes and related types of mishaps are a typical reason for injury and death among the human population. One of the elements which assume a significant part in mishaps is the human blunders including driving exhaustion depending on new brilliant methods; this application recognizes the indications of weariness and lethargy despite the individual at the hour of driving. The proposed framework depends on three separate calculations. In this model, the individual’s face is monitored by a camera in the initial step by getting 14-16fps video grouping. At that point, the pictures are changed from RGB space into YCbCr and HSV spaces. The face territory is isolated from different parts and exceptionally precise HDP is accomplished. That the eyes are open or shut in a particular time span is controlled by zeroing in on thresholding and conditions concerning the balance of human appearances. The proposed framework has been carried out in excess of thirty distinctive video successions with normal precision of93.18 and a discovery rate (DR) of 92.71 out of approximately 2500 picture outlines. High precision in the division, low blunder rate, and fast preparation of information recognize this framework from comparable ones. This framework can limit the number of mishaps brought about by drivers’ exhaustion.
This paper affords a identify and discover human faces is an photograph irrespective of their position, scale, in-aircraft rotation, orientation, pose, illumination etc. To efficaciously become aware of the individual through surprising the history and different noises within side the photograph. It have to be easy and powerful for the customers and become aware of faces which might be covered with scarf, mask, shades etc, is one of the vital components that influences the overall performance of face reputation. Many algorithms and technology are proposed to clear up the occluded face reputation wherein we use haar cascade set of rules that is simplest one. Initially this set of rules wishes loads of nice pix(pix of faces) and terrible pix(pix without faces) to teach the classifier then we should extract features from it. Facial detection is prompted through readability of the photograph, colored or black and white pix. It can simplest help frontal detection of pix and the education does takes loads of time with a purpose to separate a terrible face from a terrible face from
Currently, data pre-processing is one of the areas of nice interest as a result of it permits the discovery of hidden and infrequently attention-grabbing patterns in massive volumes of information. information scientists pay most of their time on information preparation tasks that have investigation regarding the info, loading information, and cleanup information, in line with an exploration conducted by Anaconda. The real-world massive information sets square measure obtained from several sources and contain data that tend to be incomplete, creaky, and inconsistent thence required correct investigation. during this context, it’s vital to arrange information to satisfy the necessities of information mining algorithms. this can be the role of the information pre-processing stage, within which information cleanup, transformation, and integration, or information spatiality reduction square measure performed. just about any sort of information analytics, information science or AI development needs some sort of information pre-processing to supply reliable, precise, and strong results for enterprise applications.
