Episodic self-imitation learning, a novel self-imitation algorithm with a trajectory selection module and an adaptive loss function, is proposed to speed up reinforcement learning. Compared to the original self-imitation learning algorithm, which samples good state–action pairs from the experience replay buffer, our agent leverages entire episodes with hindsight to aid self-imitation learning. A selection module is introduced to filter uninformative samples from each episode of the update. The proposed method overcomes the limitations of the standard self-imitation learning algorithm, a transitions-based method which performs poorly in handling continuous control environments with sparse rewards. From the experiments, episodic self-imitation learning is shown to perform better than baseline on-policy algorithms, achieving comparable performance to state-of-the-art off-policy algorithms in several simulated robot control tasks. The trajectory selection module is shown to prevent the agent learning undesirable hindsight experiences. With the capability of solving sparse reward problems in continuous control settings, episodic self-imitation learning has the potential to be applied to real-world problems that have continuous action spaces, such as robot guidance and manipulation.

RGB-D salient object detection (SOD) demonstrates its superiority in detecting in complex environments due to the additional depth information introduced in the data. Inevitably, an independent stream is introduced to extract features from depth images, leading to extra computation and parameters. This methodology sacrifices the model size to improve the detection accuracy which may impede the practical application of SOD problems. To tackle this dilemma, we propose a dynamic knowledge distillation (DKD) method, along with a lightweight structure, which significantly reduces the computational burden while maintaining validity. This method considers the factors of both teacher and student performance within the training stage and dynamically assigns the distillation weight instead of applying a fixed weight on the student model. We also investigate the issue of RGB-D early fusion strategy in distillation and propose a simple noise elimination method to mitigate the impact of distorted training data caused by low quality depth maps. Extensive experiments are conducted on five public datasets to demonstrate that our method can achieve competitive performance with a fast inference speed (136FPS) compared to 12 prior methods.

This paper studies the problem surrounding distributed passive arrays (sensors) locating multiple emitters while performing self-calibration to correct possible errors in the assumed array directions. In our setting, only the angle-of-arrival (AoA) information is available for localization. However, such information may contain bias due to array directional errors. Hence, localization requires self-calibration. To achieve both, the key element behind our approach is that the received signals from the same emitter should be geometrically consistent if sensor arrays are successfully calibrated. This leads to our signal model, which is built on a mapping directly from emitter locations and array directional errors to received signals. Then we formulate an atomic norm minimization and use group sparsity to promote geometric consistency and align ‘ghost’ emitter locations from calibration errors. Simulations verify the effectiveness of the proposed scheme. We derive the Cramér Rao lower bound and numerically compare it to the simulations. Furthermore, we derive a necessary condition as a rule of thumb to decide the feasibility of joint localization and calibration.