Image editing methods based on diffusion models are significantly superior to traditional methods. However, due to their slow sampling speed, high computational complexity, and weak data generalization ability, these have encountered certain limitations in practical applications. This paper proposes an efficient sparse blocks inference method for diffusion models to address this issue. It also compensates and trains the sampled feature maps by reusing low-frequency information and introduces Lp-norm to replace Euclidean distance to calculate the loss function, thereby enhancing the reconstruction effect of high-frequency image features. This method takes low-frequency sparse block data inputs as constraints, using the masks of the difference and converting them into indices to achieve the reproduction of high-resolution data. Experiments on the LSUN and CelebAHQ dataset, our method improves the inference speed of DDIM by 2.68 ×, PD by 1.53 × and SDEdit by 5.5 ×, reduces the computational complexity of DDIM by 4.1 ×, PD by 1.7 × and SDEdit by 4.8 ×.

This paper presents a low constant quiescent current and fast transient response capacitor-less low dropout regulator (CL-LDO) with high current efficiency, achieved using a dynamic slew-rate enhancement (SRE) circuit. The proposed CL-LDO achieves a fast transient response without requiring additional current under heavy loads, enabling a peak current efficiency of 99.99%. Designed using a 55-nm standard CMOS process, the CL-LDO consumes a quiescent current of 10 µA. It can deliver 0.2-100 mA of load current at a 1.0-V output from a 1.2-V to 1.5-V supply voltage. Additionally, it achieves a settling time of 0.418 µs and an overshoot/undershoot voltage of 74 mV/166 mV for load steps ranging from 0.2 mA to 100 mA, with an edge time of 0.1 µs.