Using VGG Models with Intermediate Layer Feature Maps for Static Hand Gesture Recognition

Abstract

A hand gesture recognition system provides a robust and innovative solution to nonverbal communication through human–computer interaction. Deep learning models have excellent potential for usage in recognition applications. To overcome related issues, most previous studies have proposed new model architectures or have fine-tuned pre-trained models. Furthermore, these studies relied on one standard dataset for both training and testing. Thus, the accuracy of these studies is reasonable. Unlike these works, the current study investigates two deep learning models with intermediate layers to recognize static hand gesture images. Both models were tested on different datasets, adjusted to suit the dataset, and then trained under different methods. First, the models were initialized with random weights and trained from scratch. Afterward, the pre-trained models were examined as feature extractors. Finally, the pre-trained models were fine-tuned with intermediate layers. Fine-tuning was conducted on three levels: the fifth, fourth, and third blocks, respectively. The models were evaluated through recognition experiments using hand gesture images in the Arabic sign language acquired under different conditions. This study also provides a new hand gesture image dataset used in these experiments, plus two other datasets. The experimental results indicated that the proposed models can be used with intermediate layers to recognize hand gesture images. Furthermore, the analysis of the results showed that fine-tuning the fifth and fourth blocks of these two models achieved the best accuracy results. In particular, the testing accuracies on the three datasets were 96.51%, 72.65%, and 55.62% when fine-tuning the fourth block and 96.50%, 67.03%, and 61.09% when fine-tuning the fifth block for the first model. The testing accuracy for the second model showed approximately similar results.

Composite nanostructured growth of (CdS)0.75 (PbS)0.25/Si solar cell and its characterization

“…Those figures show that at energies above the band gap, the absorption coefficient for the produced films attained greater values. This outcome is consistent with those of numerous other researchers [8][9][10][11]. According to Table 1, the absorption coefficient falls when annealing temperatures rise from (4.918 to 2.278) cm -1 , and this is because of the increasing energy gap with T a .…”

“…When films are annealed at 573K and 373K, the values of grains sizes rise from 81.24nm for films that are deposited to 90.2nm and 86.20, respectively. The production of large grains may be caused by the aggregation of atoms on the surface at higher temperatures, which is consistent with another research [8,9]. This indicates that heat treatment has increased the films' crystallinity.…”

“…These data show that the (CdS) 0.75 -(PbS) 0.25 films' absorption coefficient has been characterized by a high absorption at shorter wave-length region between 0.20 and 0.6 µm and by an absence of sharp edges on the long wavelength side from 0.64 to µ1.1 m. The coefficient of absorption shows high values at shorter wavelengths α (α > 10 4 ), indicating a high likelihood of the permitted direct transition [10-12], and subsequently falls as wavelength increases. The absorption coefficient values are almost in agreement with the figures provided by[3,8,9]. The absorption coefficient (α) of (CdS) 0.75 -(PbS) 0.25 films at various annealing temperatures (RT, 373, 473, 573K) is shown in Fig.3aand may be computed using the following equation:…”

“…It is evident that for (CdS) 0.75 -(PbS) 0.25 films as deposited, k reduces from 0.156 to 0.072 in the case where T a increases from R.T. to 573 K because T a increases diminish absorbance or the absorption coefficient, which causes k to decrease. [12] The optical energy gap values (E g ) for (CdS) 0.75 -(PbS) 0.25 films were determined with the use of the Tauc equation that is utilized for finding the optical transition type [3,[8][9][10][11][12][13]:…”

“…The decrease of E u with increased annealing temperature could be a result of the decrease of defects in the tail states and to the improvement of crystal structure by annealing [3,7,8,10] In addition to that, the optical parameters of prepared films like the refractive index (n), coefficient of extinction (k), real ((ε r ) and imaginary ((ε i ) dielectric constant parts had been calculated from transmittance and absorbance spectra, which can be seen from Table 1. Those values have been higher in the composite (CdS) 0.75 -(PbS) 0.25 films in comparison to individual CdS and PbS thin film values [8][9][10][11][12][13]. The refractive index for (CdS) 0.75 -(PbS) 0.25 films was computed from eq.5 [14]:…”