Yaprak, Büşranur; Gedikli, Eyüp

Gait, a behavior-based biometric feature, has gained increasing popularity in human identification, particularly in surveillance systems, due to its ability to function without physical contact or explicit consent. Traditional silhouette-based methods have demonstrated that different body parts exhibit distinct movement patterns during walking, thereby enhancing recognition accuracy. In this study, we propose an improved part-based gait recognition approach by leveraging ensemble learning on local body regions. The Gait Energy Image (GEI) is segmented into five horizontal parts, and ensemble learning is applied to the convolutional neural network (CNN) responsible for their processing. A separate MetaModel is trained for each body part to integrate the part-based features obtained from ensemble learning and synthesize the most discriminative ones. Additionally, a part-removal process is introduced to mitigate the effects of appearance-based variations by analyzing absolute differences between images with and without variations. The aggregated most distinctive features contribute to robust recognition. We evaluate our proposed approach on the CASIA-B, CASIA-C, and Outdoor-Gait datasets, and experimental results indicate that ensemble learning significantly enhances part-based gait recognition performance under various appearance variations, outperforming several state-of-the-art methods. The datasets and source code are available at https://github.com/busrakckugurlu/Enhancing-Part-based-Gait-Recognition-via-Ensemble-Learning-and-Feature-Fusion/tree/main.

Yaprak, Busranur; Gedikli, Eyup; Bingöl, Ozkan; Dogan, Ramazan Ozgur

Gait recognition is a biometrics-based computer vision process used to identify people based on their walking styles. Compared to other types of biometrics, gait offers a more advantageous recognition process as it does not require high-resolution and close-range images and obtains without contact. But besides this, gait biometrics is highly affected by cross-view variation, and under this variation recognition performance decreases significantly. In this study, performance evaluations of fine-tuned VGG-16 and ResNet-50 deep CNN networks on the cross-view gait recognition problem are performed. For this purpose, Gait energy images (GEI) and Silhouettes obtained from CASIA-B, the most comprehensive data set in gait recognition, are given as input to the networks. The experimental results showed that the VGG-16 network achieved higher recognition rates in cross-view gait recognition.

Yaprak, Busranur; Gedikli, Eyup

Gait recognition aims to identify people from a distance by analyzing their walking style. Nevertheless, the efficacy of recognition drops significantly under cross-view and, appearance-based variations such as carrying and clothing. In this study, the performance of the MobileNet-V1 deep network is evaluated in various scenarios to address the cross-view gait recognition problem. In the first scenario, the fine-tuned MobileNet-V1 is evaluated on Gait Energy Images (GEI) as input data, while in the second scenario, the fine-tuned MobileNet-V1 is assessed with Optical Flows and masked RGB frames input data. In the last scenario, the first two scenarios are combined over a single fused deep network based on finetuned MobileNet-V1, and a single recognition process is performed using two different fused features data; GEI features with Optical Flow features, and GEI features with masked RGB frame features. In the evaluation process, a comprehensive data set for the cross-view gait recognition problem, CASIA-B is used for the experiments. The obtained results demonstrate that in the last scenario, the contribution of masked RGB frame features to the recognition rate of GEI is more significant.

Yaprak, Büşranur; Gedikli, Eyüp

Gait recognition is the process of identifying a person from a distance based on their walking patterns. However, the recognition rate drops significantly under cross-view angle and appearance-based variations. In this study, the effectiveness of the most well-known gait representations in solving this problem is investigated based on deep learning. For this purpose, a comprehensive performance evaluation is performed by combining different modalities, including silhouettes, optical flows, and concatenated image of the Gait Energy Image (GEI) head and leg region, with GEI itself. This evaluation is carried out across different multimodal deep convolutional neural network (CNN) architectures, namely fine-tuned EfficientNet-B0, MobileNet-V1, and ConvNeXt-base models. These models are trained separately on GEIs, silhouettes, optical flows, and concatenated image of GEI head and leg regions, and then extracted GEI features are fused in pairs with other extracted modality features to find the most effective gait combination. Experimental results on the two different datasets CASIA-B and Outdoor-Gait show that the concatenated image of GEI head and leg regions significantly increased the recognition rate of the networks compared to other modalities. Moreover, this modality demonstrates greater robustness under varied carrying (BG) and clothing (CL) conditions compared to optical flows (OF) and silhouettes (SF). Codes available at https://github.com/busrakckugurlu/Different-gait-combinations-based-on-multi-modal-deep-CNN-architectures.git