Diffusion models have shown great promise in text-guided image style transfer, but there is a trade-off between style transformation and content preservation due to their stochastic nature. Existing methods require computationally expensive fine-tuning of diffusion models or additional neural network. To address this, here we propose a zero-shot contrastive loss for diffusion models that doesn’t require additional fine-tuning or auxiliary networks. By leveraging patch-wise contrastive loss between generated samples and original image embeddings in the pre-trained diffusion model, our method can generate images with the same semantic content as the source image in a zero-shot manner. Our approach outperforms existing methods while preserving content and requiring no additional training, not only for image style transfer but also for image-to-image translation and manipulation. Our experimental results validate the effectiveness of our proposed method. Code is available at https://github.com/YSerin/ZeCon.

This paper presents gain scheduling H2controller for speed control of permanent magnet synchronous motors (PMSMs) in the sense of linear parameter varying (LPV) synthesis. First, the nonlinear dynamics of a PMSM is transformed to a LPV system which has state-dependent varying parameters. The proposed method is composed of torque modulation, commutation scheme and H2controller based on parameter-dependent system. The stability of closed-loop system is proven by Lyapunov theorem. The proposed method has not only fast velocity tracking performance but also reduced energy consumption. Finally, the effectiveness of the proposed method was validated with experimental results.