This paper aims to increase the accuracy of Kazakh handwriting text recognition (KHTR) using the generative adversarial network (GAN), where a handwriting word image generator and an image quality discriminator are constructed. In order to obtain a high-quality image of handwritten text, the multiple losses are intended to encourage the generator to learn the structural properties of the texts. In this case, the quality discriminator is trained on the basis of the relativistic loss function. Based on the proposed structure, the resulting document images not only preserve texture details but also generate different writer styles, which provides better OCR performance in public databases. With a self-created dataset, images of different types of handwriting styles were obtained, which will be used when training the network. The proposed approach allows for a character error rate (CER) of 11.15% and a word error rate (WER) of 25.65%.

Ischemic stroke remains one of the leading causes of mortality and disability, and accurate segmentation of the affected areas on CT brain images plays a crucial role in timely diagnosis and clinical decision-making. This study proposes an ensemble approach based on the combination of the transformer-based models SE-UNETR and Swin UNETR using a weighted voting strategy. Its performance was evaluated using the Dice similarity coefficient, which quantifies the overlap between the predicted lesion regions and the ground-truth annotations. In this study, three-dimensional CT scans of the brain from 98 patients with a confirmed diagnosis of acute ischemic stroke were used. The data were provided by the International Tomography Center, SB RAS. The experimental results demonstrated that the ensemble based on transformer models significantly outperforms each individual model, providing more stable and accurate predictions. The final Dice coefficient reached 0.7983, indicating the high effectiveness of the proposed approach for ischemic lesion segmentation in CT images. The analysis showed more precise delineation of ischemic lesion boundaries and a reduction in segmentation errors. The proposed method can serve as an effective tool in automated stroke diagnosis systems and other applications requiring high-accuracy medical image analysis.

This paper reports a systematic study of low-rank adaptation (LoRA)-based fine-tuning applied to Qwen3 language models (4B, 8B, and 14B parameters) for the task of legal question answering within the jurisdiction of the Republic of Kazakhstan. The bilingual dataset comprises 63,114 question–answer pairs (76.2% Russian, 23.8% Kazakh) covering 11 legal domains. Models are evaluated through both automated metrics (BERTScore, citation accuracy, and hallucination rate) and blind expert assessment by a panel of two practising legal experts. Key findings: (1) all fine-tuned models reach BERTScore F1 close to 90% (89.6–90.2%) versus 82.2–83.1% for untuned base models; (2) fine-tuned models outperform GPT-4o (87.2%) and GPT-4o-mini (86.7%) on semantic similarity while exhibiting far lower hallucination rates (27–29% vs. 83–90%); (3) blind expert assessment confirms the advantage of fine-tuned models, with panel mean completeness scores of 4.28/5 versus 1.95/5 for base models (quadratically weighted Cohen’s κ = 0.80–0.95 across rating dimensions, indicating substantial to almost perfect inter-rater agreement); and (4) we identify a practical scaling plateau: paired Wilcoxon tests (n = 500) detect statistically significant but practically small differences across the 4B, 8B, and 14B fine-tuned variants (largest mean gap 0.67 pp on BERTScore F1; Cohen’s |d_z| ≤ 0.34), gains too small to justify the 3.5× parameter increase. These findings show that parameter-efficient adaptation of compact open-source models can match or exceed commercial LLMs for specialised legal QA in a low-resource bilingual context. We note one scope restriction: three domains (administrative, criminal, and housing law) are represented only in Russian, so the model is not validated for Kazakh language queries in these areas.

This paper describes the design of CVCER (Computer Vision and Cryptography Educational Robot) - the working model of the mobile robot used as a training stand for students. We describe the hardware and software of the robot. Hardware part of the robot can be used to teach the basics of robot architecture and software part can be used to teach the basics of computer vision and cryptography. The main goal of the project was to create a low cost, re-configurable and re-programmable robotic platform that could be used to teach the basics of the subjects described above and to test the algorithms written by students having internships in our institute and partner university.

Ischemic stroke is one of the leading causes of mortality and disability. Accurate segmentation of damaged regions in brain CT images is critical for timely diagnosis and clinical decision-making. In this study, an ensemble approach is proposed, combining SE-UNETR and Swin UNETR transformer models via weighted voting. The Dice coefficient was used for evaluation, measuring the overlap between predicted lesion regions and reference annotations. Unlike single-model approaches, ensemble neural network methods provide higher reliability and segmentation accuracy by integrating predictions from multiple architectures. Three-dimensional CT scans of 98 patients with acute ischemic stroke, provided by the International Tomography Center of the Siberian Branch of the Russian Academy of Sciences, were used. The results demonstrated that the proposed ensemble outperforms individual models. The average Dice coefficient was 0.7983, indicating the high effectiveness of the method in segmenting ischemic lesions. Analysis showed that the ensemble approach more accurately delineates lesion boundaries in brain CT images and reduces segmentation errors. The proposed method can be applied not only to stroke but also to other pathologies requiring precise medical image analysis in automated diagnostic systems.

Accurate and automated segmentation of ischemic stroke lesions from magnetic resonance imaging (MRI) data is crucial for rapid clinical diagnosis, treatment planning, and patient prognosis. This study presents a comprehensive evaluation of various deep learning architectures for segmenting acute and subacute ischemic lesions using the multicenter ISLES 2022 dataset. We compare traditional 3D U-Net models, attention mechanisms, and modern hybrid architectures, specifically Swin-UNETR (Swin Transformer-based UNet TRansformer). Our experiments utilize multimodal MRI data, including diffusion-weighted images (DWI) and apparent diffusion coefficient (ADC) maps. Results demonstrate that the Swin-UNETR architecture, combined with a specialized pad or crop preprocessing pipeline, advanced data augmentation using the TorchIO library, and a combined Dice–Cross Entropy (DiceCE) loss function, provides competitive performance. The best Swin-UNETR model achieved a Dice of 0.717, precision of 0.793, recall of 0.708, and HD95 of 12.24 mm on the validation set, with a loss of 0.26. The baseline 3D U-Net achieved a comparable Dice of 0.726, while Swin-UNETR showed higher precision, recall, and lower HD95. This study highlights the effectiveness of integrating hierarchical transformers with convolutional decoders for extracting both local and global contextual information in complex medical imaging tasks.