Twin delayed deep deterministic (TD3) policy gradient is an effective algorithm for continuous action spaces. However, it cannot efficiently explore the spatial space and suffers from slow convergence, which is mainly due to the serial mode strategy in learning policies. On the other hand, asynchronous reinforcement learning algorithms, e.g., asynchronous advantageous actor-critic (A3C), are effective in exploring the environment, but they ignore the gradient information of different local agents, resulting in limiting their performance. To solve the above problems, in this study, we propose an asynchronous twin delayed deep deterministic, denoted as A-TD3, algorithm with an adaptive update strategy for continuous action spaces. Specifically, a parallel mechanism is used to improve the convergence speed, along with two adaptive weight functions based on Off-policy learning to dynamically adjust the weights of the local agents. The experimental results show that the proposed A-TD3 algorithm is able to produce comparable results in terms of training execution time and convergence speed as compared with that of conventional TD3 and other state-of-the-art algorithms.

Convolutional Neural Network (CNN) achieves impressive success in image super-resolution (SR), where global context interaction is critical for reconstructing reliable edge and texture details. However, most CNN-based SR models focus on modeling the global contextual information within a single feature map by using attention mechanisms, and ignore the dependencies among hierarchical features, resulting in blurred or even distorted detail restoration, especially for SR tasks with large scaling factors (i.e., $\times 4$ , $\times 8$ ). To tackle the above issue, we propose a Cross-Layer interaction pyramid Super-Resolution (CLSR) network that reconstructs the desired SR images progressively in a coarse-to-fine fashion. Specifically, we propose a novel Cross-Layer Non-Local attention (CLNL) for accurate detail restoration. Through explicitly modeling the long-range feature-wise similarities within and between layers, the proposed CLNL is able to discriminatively explore complementary patches from hierarchical features to reconstruct the target LR patches. Then, to further strengthen the information interaction among hierarchical features at different scales, we propose a novel Gradient Consistency-Aware learning framework (GCA) by constructing a closed loop (LR $\rightarrow $ HR $\rightarrow $ LR) on the gradient space. The proposed GCA is able to effectively capture the interdependence between LR and HR gradient maps to guide our CLSR for reliable detail restoration. Extensive experiments validate that our CLSR outperforms the state-of-the-art methods in terms of both reconstruction accuracy and visual quality.

Recently, diffusion models as a hot paradigm have shown considerable superiority in image restoration with an unsupervised manner. However, they require iterative refinement from isotropic Gaussian through thousands of steps to produce a sample with exceptional quality. Most existing methods are devoted to designing fast solvers for the reverse stochastic differential equation (SDE) to accelerate sampling, while neglecting the potential of forward SDE. To better stimulate this potential, we propose the Torch-Advent-Civilization-Evolution (TACE), a novel diffusion model-based zero-shot framework for image restoration. Specifically, we propose the “Torch”, a latent vector that explicitly contains content information from the measurement image. By utilizing the Torch instead of isotropic Gaussian as initialization, our TACE significantly accelerates image restoration with better consistency and realness. To acquire the Torch from the forward process, we propose Prometheus SDEs, a cluster of equivalent SDEs. Furthermore, we construct a Conditional Guidance Projection (CGP) for the reverse SDE to strengthen the consistency of restored images. Finally, we design a Civilization Shuttle Strategy (CSS) for the generation process to enhance the realness of restored images. Extensive experiments validate that our TACE achieves state-of-the-art performance with fewer sampling steps in various typical tasks, such as super-resolution, deblurring, and colorization.

The prioritized experience replay mechanisms have achieved remarkable success in accelerating the convergence of reinforcement learning algorithms. However, applying traditional prioritized experience replay mechanisms directly to asynchronous reinforcement learning leads to slow convergence, due to the difficulty for an agent to utilize excellent experiences obtained by other agents interacting with the environment. To address the above issue, we propose a Multi-pool Prioritized experience replay-based asynchronous Twin Delayed Deep Deterministic policy gradient algorithm (MP-TD3). Specifically, a multi-pool prioritized experience replay mechanism is proposed to strengthen the experience interactions among different agents, thus speeding up the convergence. Then, a global-pool self-cleaning mechanism based on sample diversity and a global-pool self-cleaning mechanism based on TD-errors are designed to overcome the deficiency that the samples suffer from high redundancy and low information content in the global-pool, respectively. Finally, a multi-batch sampling mechanism is investigated to further reduce the training time. Extensive experiments validate that the proposed MP-TD3 significantly improve the convergence speed and performance compared with state-of-the-art methods.

Supervised Pix2Pix and unsupervised Cycle-consistency are two modes that dominate the field of medical image-to-image translation. However, neither modes are ideal. The Pix2Pix mode has excellent performance. But it requires paired and well pixel-wise aligned images, which may not always be achievable due to respiratory motion or anatomy change between times that paired images are acquired. The Cycle-consistency mode is less stringent with training data and works well on unpaired or misaligned images. But its performance may not be optimal. In order to break the dilemma of the existing modes, we propose a new unsupervised mode called RegGAN for medical image-to-image translation. It is based on the theory of "loss-correction". In RegGAN, the misaligned target images are considered as noisy labels and the generator is trained with an additional registration network to fit the misaligned noise distribution adaptively. The goal is to search for the common optimal solution to both image-to-image translation and registration tasks. We incorporated RegGAN into a few state-of-the-art image-to-image translation methods and demonstrated that RegGAN could be easily combined with these methods to improve their performances. Such as a simple CycleGAN in our mode surpasses latest NICEGAN even though using less network parameters. Based on our results, RegGAN outperformed both Pix2Pix on aligned data and Cycle-consistency on misaligned or unpaired data. RegGAN is insensitive to noises which makes it a better choice for a wide range of scenarios, especially for medical image-to-image translation tasks in which well pixel-wise aligned data are not available

Transformers have demonstrated remarkable success in image super-resolution (SR) owing to their powerful long-range dependency modeling capability. Although increasing the sliding window size of transformer-based models (e.g., SwinIR) can improve SR performance, this weakens the learning of the fine-level local features, resulting in blurry details in the reconstructed images. To address this limitation, we propose a local feature enhancement transformer for image super-resolution (LFESR) that benefits from global feature capture while enhancing local feature interaction. The basis of our LFESR is the local feature enhancement transformer (LFET), which achieves a balance between the spatial processing and channel configuration in self-attention. Our LFET contains neighborhood self-attention (NSA) and a ghost head, which can be easily applied to existing SR networks based on window self-attention. First, NSA utilizes the Hadamard operation to implement a third-order mapping to enhance local interaction, thus providing clues for high-quality image reconstruction. Next, the novel ghost head combines attention maps with static matrices to increase the channel capacity, thereby enhancing the inference capability of local features. Finally, ConvFFN is incorporated to further strengthen high-frequency detail information for reconstructed images. Extensive experiments were performed to validate the proposed LFESR, which significantly outperformed state-of-the-art methods in terms of both visual quality and quantitative metrics. Especially, the proposed LFESR exceeds SwinIR by 0.49 dB and 0.52 dB in PSNR metrics at a scaling factor of 4 on Urban100 and Manga109 datasets, respectively.