A YOLO and convolutional neural network for the detection and classification of leukocytes in leukemia

The developing of deep learning systems that used for chronic diseases diagnosing is challenge. Furthermore, the localization and identification of objects like white blood cells (WBCs) in leukemia without preprocessing or traditional hand segmentation of cells is a challenging matter due to irregular and distorted of nucleus. This paper proposed a system for computer-aided detection depend completely on deep learning with three models computer-aided detection (CAD3) to detect and classify three types of WBC which is fundamentals of leukemia diagnosing. The system used modified you only look once (YOLO v2) algorithm and convolutional neural network (CNN). The proposed system trained and evaluated on dataset created and prepared specially for the addressed problem without any traditional segmentation or preprocessing on microscopic images. The study proved that dividing of addressed problem into sub-problems will achieve better performance and accuracy. Furthermore, the results show that the CAD3 achieved an average precision (AP) up to 96% in the detection of leukocytes and accuracy 94.3% in leukocytes classification. Moreover, the CAD3 gives report contain a complete information of WBC. Finally, the CAD3 proved its efficiency on the other dataset such as acute lymphoblastic leukemia image database (ALL-IBD1) and blood cell count dataset (BCCD).

Approach to hand posture recognition based on hand shape features for human–robot interaction

Hand segmentation is the initial step for hand posture recognition. To reduce the effect of variable illumination in hand segmentation step, a new CbCr-I component Gaussian mixture model (GMM) is proposed to detect the skin region. The hand region is selected as a region of interest from the image using the skin detection technique based on the presented CbCr-I component GMM and a new adaptive threshold. A new hand shape distribution feature described in polar coordinates is proposed to extract hand contour features to solve the false recognition problem in some shape-based methods and effectively recognize the hand posture in cases when different hand postures have the same number of outstretched fingers. A multiclass support vector machine classifier is utilized to recognize the hand posture. Experiments were carried out on our data set to verify the feasibility of the proposed method. The results showed the effectiveness of the proposed approach compared with other methods.

Multi-Stroke Thai Finger-Spelling Sign Language Recognition System with Deep Learning

Sign language is a type of language for the hearing impaired that people in the general public commonly do not understand. A sign language recognition system, therefore, represents an intermediary between the two sides. As a communication tool, a multi-stroke Thai finger-spelling sign language (TFSL) recognition system featuring deep learning was developed in this study. This research uses a vision-based technique on a complex background with semantic segmentation performed with dilated convolution for hand segmentation, hand strokes separated using optical flow, and learning feature and classification done with convolution neural network (CNN). We then compared the five CNN structures that define the formats. The first format was used to set the number of filters to 64 and the size of the filter to 3 × 3 with 7 layers; the second format used 128 filters, each filter 3 × 3 in size with 7 layers; the third format used the number of filters in ascending order with 7 layers, all of which had an equal 3 × 3 filter size; the fourth format determined the number of filters in ascending order and the size of the filter based on a small size with 7 layers; the final format was a structure based on AlexNet. As a result, the average accuracy was 88.83%, 87.97%, 89.91%, 90.43%, and 92.03%, respectively. We implemented the CNN structure based on AlexNet to create models for multi-stroke TFSL recognition systems. The experiment was performed using an isolated video of 42 Thai alphabets, which are divided into three categories consisting of one stroke, two strokes, and three strokes. The results presented an 88.00% average accuracy for one stroke, 85.42% for two strokes, and 75.00% for three strokes.

A Vision-Based Framework for Spotting and Segmentation of Gesture-Based Assamese Characters Written in the Air

The task of automatic gesture spotting and segmentation is challenging for determining the meaningful gesture patterns from continuous gesture-based character sequences. This paper proposes a vision-based automatic method that handles hand gesture spotting and segmentation of gestural characters embedded in a continuous character stream simultaneously, by employing a hybrid geometrical and statistical feature set. This framework shall form an important constituent of gesture-based character recognition (GBCR) systems, which has gained tremendous demand lately as assistive aids for overcoming the restraints faced by people with physical impairments. The performance of the proposed system is validated by taking into account the vowels and numerals of Assamese vocabulary. Another attribute to this proposed system is the implementation of an effective hand segmentation module, which enables it to tackle complex background settings.