An Optimized Deep Residual Network with a Depth Concatenated Block for Handwritten Characters Classification

Even though much advancements have been achieved with regards to the recognition of handwritten characters,researchers still face difficulties with the handwritten character recognition problem,especially with the advent of new datasets like the Extended Modified National Institute of Standards and Technology dataset(EMNIST).The EMNIST dataset represents a challenge for both machine-learning and deep-learning techniques due to inter-class similarity and intra-class variability.Inter-class similarity exists because of the similarity between the shapes of certain characters in the dataset.The presence of intra-class variability is mainly due to different shapes written by different writers for the same character.In this research,we have optimized a deep residual network to achieve higher accuracy vs.the published state-of-the-art results.This approach is mainly based on the prebuilt deep residual network model ResNet18,whose architecture has been enhanced by using the optimal number of residual blocks and the optimal size of the receptive field of the first convolutional filter,the replacement of the first max-pooling filter by an average pooling filter,and the addition of a drop-out layer before the fully connected layer.A distinctive modification has been introduced by replacing the final addition layer with a depth concatenation layer,which resulted in a novel deep architecture having higher accuracy vs.the pure residual architecture.Moreover,the dataset images’sizes have been adjusted to optimize their visibility in the network.Finally,by tuning the training hyperparameters and using rotation and shear augmentations,the proposed model outperformed the state-of-the-art models by achieving average accuracies of 95.91%and 90.90%for the Letters and Balanced dataset sections,respectively.Furthermore,the average accuracies were improved to 95.9%and 91.06%for the Letters and Balanced sections,respectively,by using a group of 5 instances of the trained models and averaging the output class probabilities.

Discussion of Online Teaching Strategies for Teaching Elementary Chinese Characters to Foreigners Based on Error Analyses

The expanding role of the Chinese language in international communications has become increasingly prominent as China’s comprehensive national power continues to grow,leading to a significant rise in the number of Chinese language learners.Since online teaching is not limited by time and space,its application is widespread.For beginners in the Chinese language,the Chinese characters are both a priority and a challenge.The“Chinese Character Classification,”also known as the“Six Writings,”is the earliest systematic theory of Chinese character structures,and teaching Chinese characters in categories based on the“Chinese Character Classification”is a method that fits the cognition of beginners.In order to teach Chinese characters in a targeted approach,based on the collection and analysis of the common errors of Chinese characters among beginners,(1)this paper proposes that(a)the intuitive method can be applied to teach pictographic characters,indicative characters,and associative compound characters in online teaching;(b)the inductive-deductive method of“basic characters to new characters”can be applied for the teaching of pictophonetic characters and associative compound characters;(c)the learning of character patterns should be approached in a whole-part-whole process,while importance should be attached to the suggestion of the frequency effect with a view to facilitating the online learning of Chinese characters for beginners.The aim of this paper is to provide some practical implications for the online teaching of Chinese characters to foreigners.

Recognition results classification and post-processing methods for painted characters on billet surface

Automatic identification of characters marked on billets is very important for steelworks to achieve manu- facturing and logistics informatization management. Due to the presence of adhesions, fractures, blurs, and other problems in characters painted on billets, character recognition accuracy with machine vision is relatively low, and hardly meets practical application requirements. To make the character recognition results more reliable and accu- rate, an identification results classification and post-pro- cessing method has been proposed in this paper. By analyzing issues in the image segmentation and recognition stage, the recognition result classification model, based on character encoding rules and recognition confidence, is built, and the character recognition results can be classified as correct, suspect, or wrong. In the post-processing stage, a human-machine-cooperation mechanism with a post- processing interface is designed to eliminate error infor- mation in suspect and wrong types. The system was developed and experiments conducted with images acquired in an iron and steel factory. The results show the character recognition accuracy to be approximately 89% using the character recognizer. However, this result cannot be directly applied in information management systems. With the proposed post-processing method, a human worker will query the suspect and wrong results classified by the system, determine whether the result is correct or wrong, and then, correct the wrong result through the post-processing interface. Using this method, the character recognition accuracy ultimately improves to 99.4%. Thus, the results will be more reliable applied in a practical system.

Comparison of Morphological and Genetic Differentiations in Filial Generation of Cross Between Indica and Japonica Rice

A recombinant inbred line （RIL） population of F8 and F9 generations derived from a cross between a typical indica rice （Qishanzhan） and a typical japonica rice （Akihikari） was used to study the difference between morphological differentiation based on phenotype characters and genetic differentiation using indica and japonica specific SSR markers, and to evaluate the relationship between vascular bundle characters and morphological and genetic differentiations. The results showed that the frequency distributions of morphological and genetic differentiations were all inclined to japonica type in the filial generation. The population was more inclined to japonica type based on genetic differentiation than on morphological differentiation. The consistent degrees of classification based on the Cheng’s index, the ratio of large vascular bundle number to small vascular bundle number in panicle neck （RLSVB） and the ratio of large vascular bundle number in the second internode from the top to that in the panicle neck （RLVB） were all about 50% compared with the genetic differentiation, and the consistent degree of the total scores of the Cheng’s index combined with the vascular bundle number ratios was significantly increased to about 80% compared with the genetic differentiation. Therefore, the vascular bundle characters could be used as a helpful supplement for subspecies classification.