Heart Rate (HR) analysis has always been an area of interest not only for medical field investigators but also for researchers from variety of other fields. In the last decade, research focus has been shifted towards non-contact based systems. Heart Rate Variability (HRV) analysis can be used not only to estimate the physical characteristics but also the emotional aspect of a subject. In this paper, HR is calculated from the face of a subject extracting photoplethysmography signal using Fast Fourier Transform (FFT) and filters (band-pass). A comparison is presented between two approaches, initially taking only cheeks area as region of interest (ROI) and then repeating the process with cheeks as well as nose area as ROI. A state-of-the-art dataset COHFACE is used to measure the percentage error and root mean square error (RMSE) of the system in verification. Final results show that using only cheeks area as ROI produces better result (%error 9.19 and RMSE 8.35).
A novel real-time multimodal eye blink detection method using an amalgam of five unique weighted features extracted from the circle boundary formed from the eye landmarks is proposed.The five features, namely (Vertical Head Positioning, Orientation Factor, Proportional Ratio, Area of Intersection, and Upper Eyelid Radius), provide imperative gen (z score threshold) accurately predicting the eye status and thus the blinking status.An accurate and precise algorithm employing the five weighted features is proposed to predict eye status (open/close).One state-of-the-art dataset ZJU (eye-blink), is used to measure the performance of the method.Precision, recall, F1-score, and ROC curve measure the proposed method performance qualitatively and quantitatively.Increased accuracy (of around 97.2%) and precision (97.4%) are obtained compared to other existing unimodal approaches.The efficiency of the proposed method is shown to outperform the state-of-the-art methods.
Every week, millions of dollars are pumped into the sport of football (Association Football). The transfer value of football players proliferates each year and transfer records get shattered every other transfer window. Players are signed with a certain transfer fee which is determined during the transfer of that player. However, it is noteworthy to have a metric or economic valuation signifying the price tag on a player throughout the season rather than only during the transfer. This paper introduces the contribution of consistency, popularity, crowd estimation and performance parameters on top of the factors used in previous studies in predicting market value of the player using machine learning algorithms. The results show that the predicting accuracy is enhanced when these parameters are considered for evaluating market value of the football player.
In this paper, unique features of the segmented image samples are extracted by using two major feature extraction techniques: Local Binary Pattern (LBP) and Histogram of Oriented Gradients (HOG). After this, these features are fused to get more precise and productive outcomes. The average accuracy of the three distinct datasets that were generated using the LBP and HOG features are determined. To calculate the accuracy of the three distinct models, classification techniques like KNN and SVM, are adopted.
Image caption generation is a stimulating multimodal task. Substantial advancements have been made in thefield of deep learning notably in computer vision and natural language processing. Yet, human-generated captions are still considered better, which makes it a challenging application for interactive machine learning. In this paper, we aim to compare different transfer learning techniques and develop a novel architecture to improve image captioning accuracy. We compute image feature vectors using different state-of-the-art transferlearning models which are fed into an Encoder-Decoder network based on Stacked LSTMs with soft attention,along with embedded text to generate high accuracy captions. We have compared these models on severalbenchmark datasets based on different evaluation metrics like BLEU and METEOR.
Facial micro-expressions, as the name suggests, are very concise, unprompted facial expressions that appear either intentionally or unconsciously. These are challenging to detect, because they are of very short duration (<200 ms). It has become an interesting domain for analytical research in Machine Vision and the field of Psychology. Imitated micro-expressions are extremely hard to uncover. The aim of this research paper is to comprehensively review the existing micro-expression technological developments and compare the various data sets and give a recommended solution for future research. Challenges, such as limited data sets, overfitting, and data collection methods are discussed. We conclude by assessing the shortcomings of the existing models and suggesting possible solutions for advancing futuristic micro-expression research.
Fatigue leads to tiredness, exhaustion, and sleepiness. Driving in fatigue conditions is considered dangerous and can cause serious road accidents. According to reports about 25% of road accidents are due to driver drowsiness. The main reason behind drowsiness is fatigue. While driving continuously on long trips, drivers feel sleepy. In this paper, we proposed a novel approach that is efficient enough to detect driver drowsiness accurately. An intelligent system, that can quickly and precisely determine whether the driver is feeling drowsiness or not during driving and can also generate a warning in real-time scenarios is implemented. Thus, resulting in reducing the number of accidents that take place due to the drowsiness of the drivers as well as the death rate. In this paper, drowsiness is detected by observing facial features such as Eyes and Mouth.
Micro-expression comes under nonverbal communication, and for a matter of fact, it appears for minute fractions of a second. One cannot control micro-expression as it tells about our actual state emotionally, even if we try to hide or conceal our genuine emotions. As we know that micro-expressions are very rapid due to which it becomes challenging for any human being to detect it with bare eyes. This subtle-expression is spontaneous, and involuntary gives the emotional response. It happens when a person wants to conceal the specific emotion, but the brain is reacting appropriately to what that person is feeling then. Due to which the person displays their true feelings very briefly and later tries to make a false emotional response. Human emotions tend to last about 0.5 - 4.0 seconds, whereas micro-expression can last less than 1/2 of a second. On comparing micro-expression with regular facial expressions, it is found that for micro-expression, it is complicated to hide responses of a particular situation. Micro-expressions cannot be controlled because of the short time interval, but with a high-speed camera, we can capture one's expressions and replay them at a slow speed. Over the last ten years, researchers from all over the globe are researching automatic micro-expression recognition in the fields of computer science, security, psychology, and many more. The objective of this paper is to provide insight regarding micro-expression analysis using 3D CNN. A lot of datasets of micro-expression have been released in the last decade, we have performed this experiment on SMIC micro-expression dataset and compared the results after applying two different activation functions.
'%3e%3ccircle r='20' fill='%23008'/%3e%3ccircle r='17.5' fill='%23fff'/%3e%3ccircle r='3.5' fill='%23008'/%3e%3cg id='d'%3e%3cg id='c'%3e%3cg id='b'%3e%3cg id='a' fill='%23008'%3e%3ccircle r='.875' transform='rotate(7.5 -8.75 133.5)'/%3e%3cpath d='M0 17.5.6 7 0 2l-.6 5L0 17.5z'/%3e%3c/g%3e%3cuse xlink:href='%23a' transform='rotate(15)'/%3e%3c/g%3e%3cuse xlink:href='%23b' transform='rotate(30)'/%3e%3c/g%3e%3cuse xlink:href='%23c' transform='rotate(60)'/%3e%3c/g%3e%3cuse xlink:href='%23d' transform='rotate(120)'/%3e%3cuse xlink:href='%23d' transform='rotate(-120)'/%3e%3c/g%3e%3c/svg%3e)