The paper introduces UVM-Net, an efficient single-image dehazing network that integrates the local feature extraction ability of convolutional layers with the long-range dependency modeling capability of State Space Sequence Models (SSMs). Inspired by SSMs, the UVM-Net uses a Bi-SSM block to capture long-range dependencies effectively. The network follows an encoder-decoder structure with UVM-Net blocks in the encoder and convolution blocks in the decoder, along with skip connections. The encoder processes the input image through convolution and SSM layers, while the decoder reconstructs the dehazed image. The model is efficient and suitable for resource-constrained devices. Experimental results show that UVM-Net outperforms existing methods in image dehazing tasks. The model is also effective for other image restoration tasks. The paper also evaluates the model on low-light enhancement and deraining tasks, demonstrating its versatility. The results show that UVM-Net is a promising architecture for future image restoration networks. The code is available at https://github.com/zzr-idam. Keywords: Transformer · Image dehazing · UVM-Net · SSMs.The paper introduces UVM-Net, an efficient single-image dehazing network that integrates the local feature extraction ability of convolutional layers with the long-range dependency modeling capability of State Space Sequence Models (SSMs). Inspired by SSMs, the UVM-Net uses a Bi-SSM block to capture long-range dependencies effectively. The network follows an encoder-decoder structure with UVM-Net blocks in the encoder and convolution blocks in the decoder, along with skip connections. The encoder processes the input image through convolution and SSM layers, while the decoder reconstructs the dehazed image. The model is efficient and suitable for resource-constrained devices. Experimental results show that UVM-Net outperforms existing methods in image dehazing tasks. The model is also effective for other image restoration tasks. The paper also evaluates the model on low-light enhancement and deraining tasks, demonstrating its versatility. The results show that UVM-Net is a promising architecture for future image restoration networks. The code is available at https://github.com/zzr-idam. Keywords: Transformer · Image dehazing · UVM-Net · SSMs.