The field of signal and image enhancement is rapidly advancing with the development of new deep learning methods. Researchers are exploring the use of convolutional neural networks (CNNs) and transformers to improve the accuracy and efficiency of enhancement tasks. One notable trend is the integration of physical priors into deep neural networks, such as the use of signal-to-noise ratio (SNR) priors to reduce wavelength-dependent attenuation in underwater image enhancement. Another area of focus is the development of hybrid models that combine the strengths of different architectures, such as the use of U-Net-based encoder-decoder architectures with transformer encoders for robust ECG denoising. These innovative approaches are achieving state-of-the-art results in various applications, including underwater image enhancement and ECG denoising. Noteworthy papers include: SFormer, which proposes a novel SNR-guided transformer for underwater image enhancement, achieving a 3.1 dB gain in PSNR and 0.08 in SSIM. TF-TransUNet1D, which introduces a time-frequency guided transformer U-Net for robust ECG denoising, demonstrating consistent superiority over state-of-the-art baselines in terms of SNR improvement and error metrics.