An ideal hand prosthesis should provide satisfying functionality based on reliable decoding of the user's intentions and deliver tactile feedback in a natural manner. The absence of tactile feedback impedes the functionality and efficiency of dexterous hand prostheses, which leads to a high rejection rate from prostheses users. Thus, it is expected that integration of tactile feedback with hand prostheses will improve the manipulation performance and enhance perceptual embodiment for users. This paper reviews the state-of-the-art of non-invasive stimulation-based tactile sensation for upper-extremity prostheses, from the physiology of the human skin, to tactile sensing techniques, non-invasive tactile stimulation, and an emphasis on electrotactile feedback. The paper concludes with a detailed discussion of recent applications, challenging issues, and future developments.

Tactile feedback plays an important role in hand manipulation, especially in the grasping process which is one of the major functions of the hand. However, few commercially available prosthetic hands or hand motor function rehabilitation systems are equipped with tactile feedback. The absence of suitable tactile feedback modules leads to an inferior rehabilitation performance with a large burden on user training and compromised usability. Thus, it is challenging and essential to integrate a proper tactile feedback module with the existing hand rehabilitation systems to achieve a better control performance and accelerate the rehabilitation process. This paper focuses on the implementation and evaluation of the electrotactile feedback (EF) enhanced rehabilitation system. A virtual hand rehabilitation platform is proposed comprising an surface electromyography (sEMG) acquisition module, an electrotactile stimulation module, a virtual environment with sEMG-driven humanlike hand and numerical feedbacks of grasping force and fingertip deformation, where a closed-loop control is formed. Three different feedback conditions including visual feedback (VF), EF, and no feedback (NF) are compared based on the proposed platform. Experiments were conducted on 10 able-bodied subjects, and multiple quantitative metrics for the rehabilitation performance evaluation including training burden estimation and success rate (SR) of tasks were adopted. Results indicate that the integration of EF is helpful to both reduce the rehabilitation duration and improve the virtual grasping SR in comparison with the NF condition while possessing a better practicality over VF.

Surface electromyography (sEMG) has dominated upper-limb prosthesis control for decades due to its simplicity and effectiveness [1-3]. However, the inherent variability of EMG signal hinders the flexible and accurate control of advanced multi-functional prosthesis. This study is an attempt to use ultrasonography (US) as an alternative for prosthetic hand control. A type of multi-sensory module, comprising a single-element ultrasound channel and one sEMG bipolar channel, is customised to ensure a fair comparison between these two modalities. Three machine-learning-oriented approaches were adopted to evaluate the performance in motion classification based on datasets captured from a trans-radial amputee. The experimental results demonstrated that the ultrasound outperformed the sEMG in random (98.9% vs 70.4%) and enhanced-trial-wise (74.10% vs 61.83%) cross-validation, but fell behind the sEMG in trial-wise (39.47% vs 58.04%) validation that is the closest comparison to a real life prosthetic control. This study preliminarily implies that 1) A-mode ultrasound signal can be more stable than the sEMG with minimum electrode shift, but more sensitive to external interference than the sEMG; and 2) to maintain high classification accuracy, US approach may require harsher electrode fixing mechanism or advanced on-line calibration approach.

Accurate force control plays a crucial role but still faces many challenges in functional electrical stimulation (FES) based neuroprosthesis applications. This study proposes an surface electromyography (sEMG) driven solution for FES sequential parameter estimation including its amplitude and pulse width. The algorithm forms the mapping relationship from voluntary sEMG to the two FES parameters via the correlated muscle forces as an intermediate variable. Six able-bodied subjects were recruited to validate the proposed method based on the triangular and random grip force. The estimated amplitude and pulse width were evaluated respectively by stimulating the ipsilateral and contralateral muscle of the sEMG recording points. The accuracy of muscle force reconstruction achieved by the method were investigated by comparing the FES-induced grip force and the originally recorded voluntary grip force, in terms of the correlation index (R). The experimental results revealed excellent performance of the method on inducing functional muscle contraction closely matched the voluntary grip force (R > 0.9), regardless of the estimated amplitudes or pulse widths. Furthermore, the performance differences between the estimated pulse width and amplitude were not significant (p > 0.05). These results demonstrated the feasibility of the proposed method and its practical potential for muscle force control.

Surface electromyogram (sEMG)-based hand gesture recognition, which interprets commands given by humans through sEMG signals, performs well in many studies. However, its recognition accuracy drops dramatically due to electrode shift since the distributions of motion classes are changed. Although calibrating the system with newly collected samples after electrode shift maintains the accuracy, collecting labeled samples is inconvenient and time-consuming since the procedure is rigid. However, the calibration may not work properly without label, especially when the change is significant. This study proposes a user friendly and convenient calibration method for hand gesture recognition by an unsupervised domain adaptation method, which only obtains the unlabeled samples of preselected benchmark classes from users in calibration. The change of benchmark classes is captured by unlabeled samples by a clustering method. The other classes are estimated based on the benchmark classes by regression models. As a result, the information of all classes is used to calibrate the system. Linear discriminant analysis is used to demonstrate our model. A dataset with ten subjects is collected to verify the performance empirically. Experimental results confirm that our method utilizes the unlabeled benchmark class samples in calibration and achieves 75.55% average accuracy. Our method is more robust to electrode shift and improves around 8.5% accuracy consistently on all subjects compared with the methods without calibration or label information in calibration. Although the accuracy of our method is slightly less than the ones using label calibration samples, our calibration data collection is more convenient and less complicated.

Existing dehazing methods deal with real-world haze images with difficulty, especially scenes with thick haze. One of the main reasons is lacking real-world pair data and robust priors. To improve dehazing ability in real-world scenes, we propose a semi-supervised codebook learning dehazing method. The codebook is used as a strong prior to guide the hazy image recovery process. However, the following two issues arise when the codebook is applied to the image dehazing task: (1) Latent space features obtained from the coding of degraded hazy images suffer from matching errors when nearest-neighbour matching is performed. (2) Maintaining a good balance of image recovery quality and fidelity for heavily degraded dense hazy images is difficult. To reduce the nearest-neighbor matching error rate in the vector quantization stage of VQGAN, we designed the unit dual-attention residual transformer module (UDART) to correct the latent space features. The UDART can make the latent features obtained from the encoding stage closer to those of the corresponding clear image. To balance the quality and fidelity of the dehazing result, we design a haze density guided weight adaptive module (HDGWA), which can adaptively adjust the multi-scale skip connection weights according to haze density. In addition, we use mean teacher, a semi-supervised learning strategy, to bridge the domain gap between synthetic and real-world data and enhance the model generalization in real-world scenes. Comparative experiments show that our method achieves improvements of 0.003, 2.646, and 0.019 over the second-best method for the no-reference metrics FADE, MUSIQ, and DBCNN, respectively, on the real-world dataset URHI.