Agradezco a profesores y amigos que han compartido conmigo nuevos conocimientos y enseñanzas académicas para seguir valorando lo que hago día a día.

No abstract

Visual interpretation of sign language gesture can be useful in accomplishing natural human robot interaction. This paper describes a sign language gesture based recognition, interpreting and imitation learning system using Indian Sign Language for performing Human Robot Interaction in real time. It permits us to construct a convenient sign language gesture based communication with humanoid robot. The classification, recognition, learning, interpretation process is carried out by extracting the features from Indian sign language (ISL) gestures. Chain code and fisher score is considered as a feature vector for classification and recognition process. It is to be done by the two statistical approaches namely known as Hidden Markov Model (HMM) technique and feed forward back propagation neural network (FNN) in order to achieve satisfactory recognition accuracy. The sensitivity, specificity and accuracy were found to be equal 98.60%, 97.64% and 97.52% respectively. It can be concluded that FNN gives fast and accurate recognition and it works as promising tool for recognition and interpretation of sign language gesture for human computer interaction. The overall accuracy of recognition and interpretation of the proposed system is 95.34%. Thus, this approach is suitable for automated real time human computer interaction tool.

In this research work, to understand the types of cancer cell and attempt to analyses the biopsy slides. In this method to identify cancer parts just using simple technique of isolation of insignificant portion of biopsy slide by cancer cell level and object level segmentation and classification. Many features used in the cancer cell detection and classification of biopsy image are inspired by clinical pathologists as important for diagnosis, prognosis and characterization. A large majority of these features are features of cell nuclei in biopsy image; as such, there is often the desire to segment the image into individual cell nuclei and cancer object. In this paper, present an analysis of the utility of color Thresholding, adaptive Thresholding and watershed method for segmentation of cancer cell nuclei for classification of H&E stained histopathology image of breast tissue using neural network. This paper showing the cell level and object level classification performance using these segmented nuclei in a benign versus malignant. Results indicate that very good segmentation and classification accuracies can be achieved with color Thresholding, adaptive Thresholding, watershed based segmentation of cancer cell nuclei and cancer objects and classification of biopsy image.

This paper aims learning powerful picture portrayal for single preparing test per individual face acknowledgment. Propelled at achievement of deep learning (DL) within picture portrayal, this article proposes a regulated autoencoder (AE), which is another sort of edifice obstruct for deep structures. There are 2 highlights particular our managed AE from standard AE. In the initial place, we uphold the countenances by variations to be delineated using the standard substance of the diacritic, as example, frontispiece having unbiased appearance and typical enlightenment, further, we uphold features comparing to an analogous diacritic to be comparable. Therefore, our administered autoencoder separates the highlights which are strong to fluctuations in light, articulation, impediment, and present, and encourages the face acknowledgment. We heap such regulated AEs to achieve the deep design and utilize it for separating highlights in picture portrayal. Exploratory corollary on the Extended Yale B (Athinodoros Georghiades, et al., 2001. From few to many: Illumination cone models for face recognition under variable lighting and pose. PAMI, 23(6), pp.643–660), informational collections show that by coupling with the normally utilized inadequate portrayal based grouping, our stacked managed AEs-based face portrayal essentially beats the usually utilized picture portrayals in single example per individual face acknowledgment, and it attains lofty cognizance precision conflicted and alternative DL prototypes, which includes the deep Lambertian organize, incite of appreciably fewer preparing detail and without area data. Additionally, regulated AE can correspondingly be utilized for face check that additionally shows its acceptability for frontispiece rendering.