TENCON 2019 - 2019 IEEE Region 10 Conference (TENCON)
A gait recognition system usually degrades a lot due to intra-subject variations, such as changin... more A gait recognition system usually degrades a lot due to intra-subject variations, such as changing of views, carrying bags during gait sequences. This paper proposes an unsupervised classification algorithm based on diverse viewpoints of gait sequences with respect to normal walk and carrying conditions. This can be achieved with the help of Kernel PCA ($KPCA$) and Minimum Spanning Tree (MST) based clustering. Kernel PCA is a nonlinear form of PCA exploits the spatial structure of gait features. MST based clustering is implemented for classifying different intra subject variations into different clusters. Independent clusters are modeled for different conditions of gait sequences by using successive removal of overlapping nodes, and outliers. Discriminate clusters at the different conditions of training set makes the system more robust for indivisual recognition. A significant EER improvement is achieved using the proposed methods such as (PCA-MST) and (KPCA-MST). To evaluate the performance of the proposed method, experiments are carried out with CASIA dataset to demonstrate the efficacy of the state-of-art techniques.
Proceedings of the Twelfth Indian Conference on Computer Vision, Graphics and Image Processing, 2021
A smartphone-based gait recognition system is very interesting research in surveillance. Its goal... more A smartphone-based gait recognition system is very interesting research in surveillance. Its goal is to recognize a target user from their walking pattern using the inertial signal. However, the performance in realistic scenarios is unsatisfactory due to several covariate factors such as carrying conditions, different surface types, wearing different shoes, wearing different clothes, and also unconstrained placing of mobile phone during walking which affects gait sample data captured by sensors. Recently, many traditional single-scale CNN networks are employed for sensor-based gait recognition. However, these have limited capability to classify only normal gait samples without covariate factors. To address these challenges, in this paper, a novel discriminative Multiscale CNN network (DMSCNN) is designed to introduce both local and global feature extraction procedures for improving classification accuracy. At first, the proposed network discovers the coarse-grained features (local feature) using multiscale CNN analysis to handle different covariate-based variation effects and highlights the significance of local features with respect to class-specific samples by incorporating a class-specific weight update network. Further, fused them to get global features for improving the overall recognition rate. The experiments are performed to evaluate the robustness of the proposed model using four benchmark datasets. The result shows that the proposed model achieves higher accuracy in identification as compared to other state-of-art methods.
2017 IEEE 15th Student Conference on Research and Development (SCOReD), 2017
Computer vision applications such as object classification, human detection, action recognition, ... more Computer vision applications such as object classification, human detection, action recognition, gait recognition, etc. are often facing challenges in terms of improper segmentation and tracking due to shadow effect. However, conventional shadow detection algorithms highlight the shadow variant and invariant features. The limitation comes from the fact that many approaches are not applicable for both outdoor and indoor shadows. They fail to detect shadow in different illumination conditions as well as a different geometric position such as ground shadow, vertical shadow, self-shadow, etc. Moreover, the limitation includes shadow detection in video sequence, where different threshold values have been computed for each change of frames due to the dynamic nature of the video sequence. As a result the complexity of the system increases. To overcome the above challenges, this paper proposes a fuzzy rule based model for cast shadow and self-shadow detection using three premises, variant p...
Human gait is a new biometric resource in visual surveillance system. It can recognize individual... more Human gait is a new biometric resource in visual surveillance system. It can recognize individual as the way they walk. In the walking process, the human body shows regular periodic variation, such as upper and lower limbs, knee point, thigh point, height, etc. which reflects the individual’s unique movement pattern. However from a computational perspective, it is quite difficult to extract some feature points (knee, thigh, leg, and hip) because of occlusion of clothes, carrying bags. Height is one of the important features from the several gait features which is not influenced by the camera performance, distance and clothing style of the subject. This paper proposes DLT method of predicting height variation signal from the gait cycle of each subject. Height estimation has done using calibrated camera images. The variation of height signal is further analyzed using various transform: DHT, DFT, and DCT. Euclidian distance and MSE are computed on feature vectors to recognize individual.
Low seed ovule ratios have been observed in natural populations of Polygala vayredae Costa, a nar... more Low seed ovule ratios have been observed in natural populations of Polygala vayredae Costa, a narrowly endemic species from the oriental pre-Pyrenees. To evaluate physical and nutritional constraints and pollen tube attrition in this endemic species, stigma and style anatomy, as well as pollen tube development along the pistil were investigated using light and fluorescence microscopy. The structural morphology of the stigmatic region was also examined with scanning electron microscopy. Pollen grains that reached the stigmatic papillae came into contact with a lipid-rich exudate and germinated easily. Although a large number of pollen grains reach the stigmatic papillae, few pollen tubes were able to grow into the style towards the ovary. The style was hollow, with the stylar channel beginning a few cells below the stigmatic papillae. Initially, the stylar channel area was small compared to other levels of the style, and was surrounded by lipid-rich, highly metabolic active cells. Furthermore, lipid-rich mucilage was detected inside the stylar channel. At subsequent style levels towards the ovary, no major reserves were detected histochemically. The reduced intercellular spaces below the stigmatic papillae and the reduced area of the stylar channel at its commencement are suggested to physically constrain the number of pollen tubes that can develop. In subsequent levels of the style, the stylar channel could physically support a larger number of pollen tubes, but the lack of nutritional reserves cannot be disregarded as a cause of pollen tube attrition. Finally, the number of pollen tubes entering the ovary was greater than the number of ovules, suggesting that interactions occurring at this level play a major role in the final reproductive outcome in this species.
TENCON 2019 - 2019 IEEE Region 10 Conference (TENCON)
A gait recognition system usually degrades a lot due to intra-subject variations, such as changin... more A gait recognition system usually degrades a lot due to intra-subject variations, such as changing of views, carrying bags during gait sequences. This paper proposes an unsupervised classification algorithm based on diverse viewpoints of gait sequences with respect to normal walk and carrying conditions. This can be achieved with the help of Kernel PCA ($KPCA$) and Minimum Spanning Tree (MST) based clustering. Kernel PCA is a nonlinear form of PCA exploits the spatial structure of gait features. MST based clustering is implemented for classifying different intra subject variations into different clusters. Independent clusters are modeled for different conditions of gait sequences by using successive removal of overlapping nodes, and outliers. Discriminate clusters at the different conditions of training set makes the system more robust for indivisual recognition. A significant EER improvement is achieved using the proposed methods such as (PCA-MST) and (KPCA-MST). To evaluate the performance of the proposed method, experiments are carried out with CASIA dataset to demonstrate the efficacy of the state-of-art techniques.
Proceedings of the Twelfth Indian Conference on Computer Vision, Graphics and Image Processing, 2021
A smartphone-based gait recognition system is very interesting research in surveillance. Its goal... more A smartphone-based gait recognition system is very interesting research in surveillance. Its goal is to recognize a target user from their walking pattern using the inertial signal. However, the performance in realistic scenarios is unsatisfactory due to several covariate factors such as carrying conditions, different surface types, wearing different shoes, wearing different clothes, and also unconstrained placing of mobile phone during walking which affects gait sample data captured by sensors. Recently, many traditional single-scale CNN networks are employed for sensor-based gait recognition. However, these have limited capability to classify only normal gait samples without covariate factors. To address these challenges, in this paper, a novel discriminative Multiscale CNN network (DMSCNN) is designed to introduce both local and global feature extraction procedures for improving classification accuracy. At first, the proposed network discovers the coarse-grained features (local feature) using multiscale CNN analysis to handle different covariate-based variation effects and highlights the significance of local features with respect to class-specific samples by incorporating a class-specific weight update network. Further, fused them to get global features for improving the overall recognition rate. The experiments are performed to evaluate the robustness of the proposed model using four benchmark datasets. The result shows that the proposed model achieves higher accuracy in identification as compared to other state-of-art methods.
2017 IEEE 15th Student Conference on Research and Development (SCOReD), 2017
Computer vision applications such as object classification, human detection, action recognition, ... more Computer vision applications such as object classification, human detection, action recognition, gait recognition, etc. are often facing challenges in terms of improper segmentation and tracking due to shadow effect. However, conventional shadow detection algorithms highlight the shadow variant and invariant features. The limitation comes from the fact that many approaches are not applicable for both outdoor and indoor shadows. They fail to detect shadow in different illumination conditions as well as a different geometric position such as ground shadow, vertical shadow, self-shadow, etc. Moreover, the limitation includes shadow detection in video sequence, where different threshold values have been computed for each change of frames due to the dynamic nature of the video sequence. As a result the complexity of the system increases. To overcome the above challenges, this paper proposes a fuzzy rule based model for cast shadow and self-shadow detection using three premises, variant p...
Human gait is a new biometric resource in visual surveillance system. It can recognize individual... more Human gait is a new biometric resource in visual surveillance system. It can recognize individual as the way they walk. In the walking process, the human body shows regular periodic variation, such as upper and lower limbs, knee point, thigh point, height, etc. which reflects the individual’s unique movement pattern. However from a computational perspective, it is quite difficult to extract some feature points (knee, thigh, leg, and hip) because of occlusion of clothes, carrying bags. Height is one of the important features from the several gait features which is not influenced by the camera performance, distance and clothing style of the subject. This paper proposes DLT method of predicting height variation signal from the gait cycle of each subject. Height estimation has done using calibrated camera images. The variation of height signal is further analyzed using various transform: DHT, DFT, and DCT. Euclidian distance and MSE are computed on feature vectors to recognize individual.
Low seed ovule ratios have been observed in natural populations of Polygala vayredae Costa, a nar... more Low seed ovule ratios have been observed in natural populations of Polygala vayredae Costa, a narrowly endemic species from the oriental pre-Pyrenees. To evaluate physical and nutritional constraints and pollen tube attrition in this endemic species, stigma and style anatomy, as well as pollen tube development along the pistil were investigated using light and fluorescence microscopy. The structural morphology of the stigmatic region was also examined with scanning electron microscopy. Pollen grains that reached the stigmatic papillae came into contact with a lipid-rich exudate and germinated easily. Although a large number of pollen grains reach the stigmatic papillae, few pollen tubes were able to grow into the style towards the ovary. The style was hollow, with the stylar channel beginning a few cells below the stigmatic papillae. Initially, the stylar channel area was small compared to other levels of the style, and was surrounded by lipid-rich, highly metabolic active cells. Furthermore, lipid-rich mucilage was detected inside the stylar channel. At subsequent style levels towards the ovary, no major reserves were detected histochemically. The reduced intercellular spaces below the stigmatic papillae and the reduced area of the stylar channel at its commencement are suggested to physically constrain the number of pollen tubes that can develop. In subsequent levels of the style, the stylar channel could physically support a larger number of pollen tubes, but the lack of nutritional reserves cannot be disregarded as a cause of pollen tube attrition. Finally, the number of pollen tubes entering the ovary was greater than the number of ovules, suggesting that interactions occurring at this level play a major role in the final reproductive outcome in this species.
Uploads
Papers by Sonia Das