Journal: Scientific Reports

Article Title: Reconstructing hand gestures with synergies extracted from dance movements

doi: 10.1038/s41598-025-25563-7

Figure Lengend Snippet: ( a ) Twenty-one hand-knuckle landmarks obtained from MediaPipe hand landmark model, and their corresponding anatomical areas are illustrated on the left side image. Each dot represents the joints, and line represents the Euclidean distance from each joint. Corresponding labels in the illustration are - 1. Wrist, 2. CMC-T, 3. MCP-T, 4. IP-T, 5. TIP-T, 6. MCP-I, 7. PIP-I, 8. DIP-I, 9. TIP-I, 10. MCP-M, 11. PIP-M, 12. DIP-M, 13. TIP-M, 14. MCP-R, 15. PIP-R, 16. DIP-R, 17. TIP-R, 18. MCP-P, 19. PIP-P, 20. DIP-P, 21. TIP-P. The right-side image represents the virtual hand model with 14 joints, labeled as 1. MCP-T, 2. IP-T, 3. MCP-I, 4. PIP-I, 5. DIP-I, 6. MCP-M, 7. PIP-M, 8. DIP-M, 9. MCP-R, 10. PIP-R, 11. DIP-R, 12. MCP-P, 13. PIP-P, 14. DIP-P where T-Thumb, I-Index, M-Middle, R-Ring and P-Pinky fingers. ( b ) The first five hand gestures recognized through MediaPipe from mudras and natural grasps dataset is illustrated here. Their corresponding virtual hand model obtained by applying Gaussian function to 14 joint angular velocities is shown below them. Here M – Mudras, VM – Virtual Mudras, N – Natural Grasps, VN – Virtual Natural Grasps. The numbers indicates the corresponding hand gestures in Table . Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Article Snippet: Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Techniques: Labeling

Journal: Scientific Reports

Article Title: Reconstructing hand gestures with synergies extracted from dance movements

doi: 10.1038/s41598-025-25563-7

Figure Lengend Snippet: Wrist re-referencing of the hand mudra chatura . (a) Original hand gesture as performed. (b) Re-referenced end mudra of chatura relative to the initial reference posture. (c) Virtual hand representation of chatura using 14 Gaussian profiles to model the joint angular velocities. Here, M - Mudras, VM - Virtual Mudras. The number represents the hand gesture in Table . Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Article Snippet: Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Techniques:

Journal: Scientific Reports

Article Title: Reconstructing hand gestures with synergies extracted from dance movements

doi: 10.1038/s41598-025-25563-7

Figure Lengend Snippet: The first six synergies of mudras (left to right) and first six natural synergies (left to right) are illustrated here. These synergies were extracted by applying dimensionality reduction on the angular velocity matrix of mudras and natural grasps datasets. The joint angular velocity profiles of 14 joints were then mapped to the virtual hand model, enabling a structured representation of the extracted synergies. Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Article Snippet: Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Techniques:

Journal: Scientific Reports

Article Title: Reconstructing hand gestures with synergies extracted from dance movements

doi: 10.1038/s41598-025-25563-7

Figure Lengend Snippet: The projection of joint angular velocities for the end posture of various hand gestures on to the first three principal components in illustrated here. The first row depicts the projections of Chandrakala and Mrugasheersha from the Mudras. Similarly, the second row presents the projections of Index Finger Extension and Ring gestures from the natural grasps. Moreover, these figures visualize signal strength, indicating the extent of flexion required to achieve each posture — yellow represents maximum strength, while navy blue denotes zero strength. Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Article Snippet: Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Techniques:

Journal: Scientific Reports

Article Title: Reconstructing hand gestures with synergies extracted from dance movements

doi: 10.1038/s41598-025-25563-7

Figure Lengend Snippet: Fifteen ASL hand gestures represented on the virtual hand model (leftmost column of each set) is illustrated here. Reconstructions using synergies of mudras are shown in blue (middle), while reconstructions using natural synergies are displayed on the right. Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Article Snippet: Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Techniques:

Journal: Scientific Reports

Article Title: Reconstructing hand gestures with synergies extracted from dance movements

doi: 10.1038/s41598-025-25563-7

Figure Lengend Snippet: The reconstructed end postures of three randomly selected hand gestures from each dataset, mapped to Mitra, are illustrated here. The first row showcases the reconstruction of X, Kartari Mukha, and Medium Wrap of ASL, mudras, and natural grasps dataset. In contrast, the second row represents the same hand gestures using natural synergies. Here, VM represents virtual mudras, VN represents virtual natural grasp. Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Article Snippet: Virtual 3D Hand Model was developed using V-Realm Builder, MATLAB Simulink 3D Animation , Version R2023a .

Techniques:

Journal: Sensors (Basel, Switzerland)

Article Title: Resource-Efficient FPGA Architecture for Real-Time RFI Mitigation in Interferometric Radiometers

doi: 10.3390/s24248001

Figure Lengend Snippet: Comparison of the Simulink OR Mask ( left ) with the full-precision MATLAB reference ( right ).

Article Snippet: The simulation was performed either by MATLAB R2022a HDL Coder or MATLAB R2022a Simulink (MathWorks), and the results were compared numerically to the outputs of the original MATLAB algorithm, which uses floating-point double-precision data types.

Techniques: Comparison

Journal: Sensors (Basel, Switzerland)

Article Title: Resource-Efficient FPGA Architecture for Real-Time RFI Mitigation in Interferometric Radiometers

doi: 10.3390/s24248001

Figure Lengend Snippet: Comparison of the Simulink AND Mask ( left ) with the full-precision MATLAB reference ( right ).

Article Snippet: The simulation was performed either by MATLAB R2022a HDL Coder or MATLAB R2022a Simulink (MathWorks), and the results were compared numerically to the outputs of the original MATLAB algorithm, which uses floating-point double-precision data types.

Techniques: Comparison

Journal: Sensors (Basel, Switzerland)

Article Title: Resource-Efficient FPGA Architecture for Real-Time RFI Mitigation in Interferometric Radiometers

doi: 10.3390/s24248001

Figure Lengend Snippet: Comparison of the Simulink OR Mask ( left ) with the full-precision MATLAB reference ( right ).

Article Snippet: The simulation was performed either by MATLAB R2022a HDL Coder or MATLAB R2022a Simulink (MathWorks), and the results were compared numerically to the outputs of the original MATLAB algorithm, which uses floating-point double-precision data types.

Techniques: Comparison

Journal: Sensors (Basel, Switzerland)

Article Title: Resource-Efficient FPGA Architecture for Real-Time RFI Mitigation in Interferometric Radiometers

doi: 10.3390/s24248001

Figure Lengend Snippet: Comparison of the Simulink AND Mask ( left ) with the full-precision MATLAB reference ( right ).

Article Snippet: The simulation was performed either by MATLAB R2022a HDL Coder or MATLAB R2022a Simulink (MathWorks), and the results were compared numerically to the outputs of the original MATLAB algorithm, which uses floating-point double-precision data types.

Techniques: Comparison