Hand gesture recognition for human computer interaction : a comparative study of different image features

Hand gesture recognition for human computer interaction, being a natural way of human computer interaction, is an area of active research in computer vision and machine learning. This is an area with many different possible applications, giving users a simpler and more natural way to communicate with robots/systems interfaces, without the need for extra devices. So, the primary goal of gesture recognition research is to create systems, which can identify specific human gestures and use them to convey information or for device control. For that, vision-based hand gesture interfaces require fast and extremely robust hand detection, and gesture recognition in real time. In this study we try to identify hand features that, isolated, respond better in various situations in human-computer interaction. The extracted features are used to train a set of classifiers with the help of RapidMiner in order to find the best learner. A dataset with our own gesture vocabulary consisted of 10 gestures, recorded from 20 users was created for later processing. Experimental results show that the radial signature and the centroid distance are the features that when used separately obtain better results, with an accuracy of 91% and 90,1% respectively obtained with a Neural Network classifier. These to methods have also the advantage of being simple in terms of computational complexity, which make them good candidates for real-time hand gesture recognition.

Absence of magnetically induced fractional quantization in atomic contacts

Using the mechanically controlled break junction technique at low temperatures and under cryogenic vacuum conditions we have studied atomic contacts of several magnetic (Fe, Co, and Ni) and nonmagnetic (Pt) metals, which recently were claimed to show fractional conductance quantization. In the case of pure metals we see no quantization of the conductance nor half quantization, even when high magnetic fields are applied. On the other hand, features in the conductance similar to (fractional) quantization are observed when the contact is exposed to gas molecules. Furthermore, the absence of fractional quantization when the contact is bridged by H2 indicates the current is never fully polarized for the metals studied here. Our results are in agreement with recent model calculations.

Hand detection in cluttered scene images using Fourier-Mellin invariant features

This paper proposes an automatic hand detection system that combines the Fourier-Mellin Transform along with other computer vision techniques to achieve hand detection in cluttered scene color images. The proposed system uses the Fourier-Mellin Transform as an invariant feature extractor to perform RST invariant hand detection. In a first stage of the system a simple non-adaptive skin color-based image segmentation and an interest point detector based on corners are used in order to identify regions of interest that contains possible matches. A sliding window algorithm is then used to scan the image at different scales performing the FMT calculations only in the previously detected regions of interest and comparing the extracted FM descriptor of the windows with a hand descriptors database obtained from a train image set. The results of the performed experiments suggest the use of Fourier-Mellin invariant features as a promising approach for automatic hand detection.

Hand detection in cluttered scene images using Fourier-Mellin invariant features

This paper proposes an automatic hand detection system that combines the Fourier-Mellin Transform along with other computer vision techniques to achieve hand detection in cluttered scene color images. The proposed system uses the Fourier-Mellin Transform as an invariant feature extractor to perform RST invariant hand detection. In a first stage of the system a simple non-adaptive skin color-based image segmentation and an interest point detector based on corners are used in order to identify regions of interest that contains possible matches. A sliding window algorithm is then used to scan the image at different scales performing the FMT calculations only in the previously detected regions of interest and comparing the extracted FM descriptor of the windows with a hand descriptors database obtained from a train image set. The results of the performed experiments suggest the use of Fourier-Mellin invariant features as a promising approach for automatic hand detection.

Hand gesture recognition based on morphologic features

Comunicación presentada en el IX Simposium Nacional de Reconocimiento de Formas y Análisis de Imágenes, Benicàssim, Mayo, 2001.

Hand detection on images based on deformable part models and additional features

Hand detection on images has important applications on person activities recognition. This thesis focuses on PASCAL Visual Object Classes (VOC) system for hand detection. VOC has become a popular system for object detection, based on twenty common objects, and has been released with a successful deformable parts model in VOC2007. A hand detection on an image is made when the system gets a bounding box which overlaps with at least 50% of any ground truth bounding box for a hand on the image. The initial average precision of this detector is around 0.215 compared with a state-of-art of 0.104; however, color and frequency features for detected bounding boxes contain important information for re-scoring, and the average precision can be improved to 0.218 with these features. Results show that these features help on getting higher precision for low recall, even though the average precision is similar.

Comparative study of human age estimation based on hand-crafted and deep face features

In the past few years, human facial age estimation has drawn a lot of attention in the computer vision and pattern recognition communities because of its important applications in age-based image retrieval, security control and surveillance, biomet- rics, human-computer interaction (HCI) and social robotics. In connection with these investigations, estimating the age of a person from the numerical analysis of his/her face image is a relatively new topic. Also, in problems such as Image Classification the Deep Neural Networks have given the best results in some areas including age estimation. In this work we use three hand-crafted features as well as five deep features that can be obtained from pre-trained deep convolutional neural networks. We do a comparative study of the obtained age estimation results with these features.

Association between intratumoral lymphatic microvessel density (LMVD) and clinicopathologic features in endometrial cancer: a retrospective cohort study

Background: Lymph node metastasis in endometrial cancer significantly decreases survival rate. Few data on the influence of intratumoral lymphatic microvessel density (LMVD) on survival in endometrial cancer are available. Our aim was to assess the intratumoral LMVD of endometrial carcinomas and to investigate its association with classical pathological factors, lymph node metastasis and survival. Methods: Fifty-seven patients with endometrial carcinoma diagnosed between 2000 and 2008 underwent complete surgical staging and evaluation of intratumoral LMVD and other histologic variables. Lymphatic microvessels were identified by immunohistochemical staining using monoclonal antibody against human podoplanin (clone D2-40) and evaluated by counting the number of immunostained lymphatic vessels in 10 hot spot areas at 400x magnification. The LMVD was expressed by the mean number of vessels in these 10 hot spot microscopic fields. We next investigated the association of LMVD with the clinicopathologic findings and prognosis. Results: The mean number of lymphatic vessels counted in all cases ranged between 0 and 4.7. The median value of mean LMVD was 0.5, and defined the cut-off for low and high LMVD. We identified low intratumoral LMVD in 27 (47.4%) patients and high LMVD in 30 (52.6%) patients. High intratumoral LMVD was associated with lesser miometrial and adnaexal infiltration, lesser cervical and peritoneal involvement, and fewer fatal cases. Although there was lower lymph node involvement among cases with high LMVD, the difference did not reach significance. No association was seen between LMVD and FIGO staging, histological type, or vascular invasion. On the other hand, low intratumoral LMVD was associated with poor outcome. Seventy-five percent of deaths occurred in patients with low intratumoral LMVD. Conclusion: Our results show association of high intratumoral LMVD with features related to more localized disease and better outcome. We discuss the role of lymphangiogenesis as an early event in the endometrial carcinogenesis.

Hand-held robotic device for laparoscopic surgery and training

Laparoscopic surgery (LS) has revolutionized traditional surgical techniques introducing minimally invasive procedures for diagnosis and local therapies. LSs have undeniable advantages, such as small patient incisions, reduced postoperative pain and faster recovery. On the other hand, restricted vision of the anatomical target, difficult handling of the surgical instruments, restricted mobility inside the human body, need of dexterity to hand-eye coordination and inadequate and non-ergonomic surgical instruments may restrict LS only to more specialized surgeons. To overcome the referred limitations, this work presents a new robotic surgical handheld system – the EndoRobot. The EndoRobot was designed to be used in clinical practice or even as a surgical simulator. It integrates an electromechanical system with 3 degrees of freedom. Each degree can be manipulated independently and combined with different levels of sensitivity allowing fast and slow movements. As other features, the EndoRobot has battery power or external power supply, enables the use of bipolar radiofrequency to prevent bleeding while cutting and allows plug-and-play of the laparoscopic forceps for rapid exchange. As a surgical simulator, the system was also instrumented to measure and transmit, in real time, its position and orientation for a training software able to monitor and assist the trainee’s surgical movements.

The Last Fifteen Years of Schistosomiasis in Venezuela: Features and Evolution

Control of schistosomiasis in Venezuela has been a topic of major interest and controversy among the metaxenic parasitosis. A small area of transmission of approximately 15,000 km2 was thought to be eradicated some years ago. However, some epidemiological characteristics of our transmission area have limited the success on the way toward eradication. Since 1945, when the Schistosomiasis Control Program started, the prevalence in the endemic area has decreased from 14% in 1943 to 1.4% in 1996. Until 1982, the surveillance of active cases was based on massive stool examination. Since then, the Schistosomiasis Research Group (SRG) recommended the additional use of serologic tests in the Control Program and the selective or massive chemotherapy depending on serological and parasitological prevalence of each community. At present, the real prevalence is underestimated due to the fact that approximately 80% of the individuals eliminate less than 100 eggs/g of feces. Those persons could be responsible for the maintenance of the foci going on and therefore limiting the impact of the control measures. Efforts of the SRG are being oriented toward improvement of immunodiagnostic tests by using defined antigens (enzymes) and chemically synthesized peptides, derived from relevant molecules of the parasite, either for antibodies or antigens search. On the other hand, introduction of snail competitors has been a biological weapon in the control of the intermediate host in certain areas. However, the recent reinfestation of water courses by Biomphalaria glabrata, the increased prevalence in some areas, together with important administrative changes at the Control Program of the Minister of Health, have arisen new questions and doubts, challenging the eradication strategy proposed during the last decade.

HIV-associated neurocognitive disorders (HAND): a changing pattern

Combination antiretroviral therapy has dramatically decreased the incidence of HIV-related mortality and serious opportunistic diseases, among which is HIV- associated dementia. However, minor forms of cognitive dysfunction have not disappeared and may even have increased in frequency. Aging of HIV+ patients, insufficient penetration of antiretroviral drugs into the brain with continuous low- grade viral production and inflammation may play a role. A putative neurotoxicity of combination antiretroviral therapy is controversial. In this article, we will discuss these aspects, as well as clinical and pathophysiological features shared by HIV-associated neurocognitive disorders and other neurodegenerative diseases, especially Alzheimer's disease. This article will briefly summarize the current clinical trials on neuroprotective agents, and the management of patients with neurocognitive disorders will be discussed

Constitutively active alpha-1b adrenergic receptor mutants display different phosphorylation and internalization features.

We compared the phosphorylation and internalization properties of constitutively active alpha-1b adrenergic receptor (AR) mutants carrying mutations in two distant receptor domains, i.e., at A293 in the distal part of the third intracellular loop and at D142 of the DRY motif lying at the end of the third transmembrane domain. For the A293E and A293I mutants the levels of agonist-independent phosphorylation were 150% and 50% higher than those of the wild-type alpha-1b AR, respectively. On the other hand, for the constitutively active D142A and D142T mutants, the basal levels of phosphorylation were similar to those of the wild-type alpha-1b AR and did not appear to be further stimulated by epinephrine. Overexpression of the guanyl nucleotide binding regulatory protein-coupled receptor kinase GRK2 further increases the basal phosphorylation of the A293E mutant, but not that of D142A mutant. Both the wild-type alpha-1b AR and the A293E mutant could undergo beta-arrestin-mediated internalization. The epinephrine-induced internalization of the constitutively active A293E mutant was significantly higher than that of the wild-type alpha-1b AR. In contrast, the D142A mutant was impaired in its ability to interact with beta-arrestin and to undergo agonist-induced internalization. Interestingly, a double mutant A293E/D142A retained very high constitutive activity and regulatory properties of both the A293E and D142A receptors. These findings demonstrate that two constitutively activating mutations occurring in distant receptor domains of the alpha-1b AR have divergent effects on the regulatory properties of the receptor.

Re-mapping the molecular features of the human immunodeficiency virus type 1 and human T-cell lymphotropic virus type 1 Brazilian sequences using a bioinformatics unit established in Salvador, Bahia, Brazil, to give support to the viral epidemiology studies

The analysis of genetic data for human immunodeficiency virus type 1 (HIV-1) and human T-cell lymphotropic virus type 1 (HTLV-1) is essential to improve treatment and public health strategies as well as to select strains for vaccine programs. However, the analysis of large quantities of genetic data requires collaborative efforts in bioinformatics, computer biology, molecular biology, evolution, and medical science. The objective of this study was to review and improve the molecular epidemiology of HIV-1 and HTLV-1 viruses isolated in Brazil using bioinformatic tools available in the Laboratório Avançado de Sáude Pública (Lasp) bioinformatics unit. The analysis of HIV-1 isolates confirmed a heterogeneous distribution of the viral genotypes circulating in the country. The Brazilian HIV-1 epidemic is characterized by the presence of multiple subtypes (B, F1, C) and B/F1 recombinant virus while, on the other hand, most of the HTLV-1 sequences were classified as Transcontinental subgroup of the Cosmopolitan subtype. Despite the high variation among HIV-1 subtypes, protein glycosylation and phosphorylation domains were conserved in the pol, gag, and env genes of the Brazilian HIV-1 strains suggesting constraints in the HIV-1 evolution process. As expected, the functional protein sites were highly conservative in the HTLV-1 env gene sequences. Furthermore, the presence of these functional sites in HIV-1 and HTLV-1 strains could help in the development of vaccines that pre-empt the viral escape process.

Revisit of multiple epiphyseal dysplasia: ethnic difference in genotypes and comparison of radiographic features linked to the COMP and MATN3 genes.

Multiple epiphyseal dysplasia (MED) is a genetically heterogeneous group of diseases characterized by variable degrees of epiphyseal abnormality primarily involving the hip and knee joints. The purpose of this study was to investigate the frequency of mutations in individuals with a clinical and radiographic diagnosis of MED and to test the hypothesis that characteristic radiological findings may be helpful in predicting the gene responsible. The radiographs of 74 Korean patients were evaluated by a panel of skeletal dysplasia experts. Six genes known to be associated with MED (COMP, MATN3, COL9A1, COL9A2, COL9A3, and DTDST) were screened by sequencing. Mutations were found in 55 of the 63 patients (87%). MATN3 mutations were found in 30 patients (55%), followed by COMP mutations in 23 (41%), and COL9A2 and DTDST mutations in one patient (2%) each. Comparisons of radiographic findings in patients with COMP and MATN3 mutations showed that albeit marked abnormalities in hip and knee joints were observed in both groups, the degree of involvement and the morphology of dysplastic epiphyses differed markedly. The contour of the pelvic acetabulum, the presence of metaphyseal vertical striations, and/or the brachydactyly of the hand were also found to be highly correlated with the genotypes. The study confirms that MATN3 and COMP are the genes most frequently responsible for MED and that subtle radiographic signs may give precious indications on which gene(s) should be prioritized for mutational screening in a given individual.

Individual difference factors and men's attraction to female body features /

Existing research on attraction to body features has suggested that men show general preferences for women with lower waist-to-hip ratios (WHR), larger breasts, and slender body weights. The present study intended to expand on this research by investigating several individual difference factors and their potential contribution to variation in what men find attractive in female body features. Two hundred and seventy-three men were assessed for sex-role identity, 2D:4D digit ratios (a possible marker of prenatal exposure to androgens, and thus masculinization), physical attractiveness, early sexual experiences (as indices of early sexual conditioning), and early family attitudes toward body features, as well as their current preferences for WHR, breast size, weight, and height in women. For WHR, as predicted, physical attractiveness, early sexual experiences, and lower (more masculine) right-hand 2D:4D ratios significantly predicted current preferences for more feminine (lower) WHR. Early sexual experiences significantly predicted later preferences for breast size; in addition, more masculine occupational preferences and lower (more masculine) left-hand 2D:4D ratios predicted preferences for larger breasts. Participants' height, education level, Unmitigated Agency (masculinity) scores, and early sexual experiences significantly predicted current preferences for height. Finally, early sexual experiences significantly predicted current preferences for weight. The results suggest that variation in preferences for women's bodily features can be uniquely accounted for by a number of individual difference factors. Strengths and weaknesses of the study, along with implications for future research, are discussed.