Xin Min

Haoyu Zheng, Wenqiao Zhang, Yaoke Wang et al.

Revolutionary advancements in text-to-image models have unlocked new dimensions for sophisticated content creation, such as text-conditioned image editing, enabling the modification of existing images based on textual guidance. This capability allows for the generation of diverse images that convey highly complex visual concepts. However, existing methods primarily focus on generating new images from text-image pairs and struggle to produce fine-grained animations from existing images and textual guidance without fine-tuning. In this paper, we introduce LASER, a tuning-free LLM-driven attention control framework that follows a progressive process: LLM planning, feature-attention injection, and stable animation generation. LASER leverages a large language model (LLM) to refine general descriptions into fine-grained prompts, guiding pre-trained text-to-image models to generate aligned keyframes with subtle variations. The LLM also generates control signals for feature and attention injections, enabling seamless text-guided image morphing for various transformations without additional fine-tuning. By using the same initial noise inversion from the input image, LASER receives LLM-controlled injections during denoising and leverages interpolated text embeddings to produce a series of coherent animation frames. We propose a Text-conditioned Image-to-Animation Benchmark to validate the effectiveness and efficacy of LASER. Extensive experiments demonstrate that LASER achieves impressive results in consistent and efficient animation generation, establishing it as a powerful tool for producing detailed animations and opening new avenues in digital content creation.

Guanghao Yin, Shouqian Sun, Chao Li et al.

In real-world single image super-resolution (SISR) task, the low-resolution image suffers more complicated degradations, not only downsampled by unknown kernels. However, existing SISR methods are generally studied with the synthetic low-resolution generation such as bicubic interpolation (BI), which greatly limits their performance. Recently, some researchers investigate real-world SISR from the perspective of the camera and smartphone. However, except the acquisition equipment, the display device also involves more complicated degradations. In this paper, we focus on the camera-screen degradation and build a real-world dataset (Cam-ScreenSR), where HR images are original ground truths from the previous DIV2K dataset and corresponding LR images are camera-captured versions of HRs displayed on the screen. We conduct extensive experiments to demonstrate that involving more real degradations is positive to improve the generalization of SISR models. Moreover, we propose a joint two-stage model. Firstly, the downsampling degradation GAN(DD-GAN) is trained to model the degradation and produces more various of LR images, which is validated to be efficient for data augmentation. Then the dual residual channel attention network (DuRCAN) learns to recover the SR image. The weighted combination of L1 loss and proposed Laplacian loss are applied to sharpen the high-frequency edges. Extensive experimental results in both typical synthetic and complicated real-world degradations validate the proposed method outperforms than existing SOTA models with less parameters, faster speed and better visual results. Moreover, in real captured photographs, our model also delivers best visual quality with sharper edge, less artifacts, especially appropriate color enhancement, which has not been accomplished by previous methods.