Abstract: Recently, transformers have garnered significant attention due to their exceptional capability to capture long-range dependencies in data. A critical factor contributing to their superior ...
Abstract: We propose reparameterized refocusing convolution (RefConv) as a replacement for regular convolutional layers, which is a plug-and-play module to improve the performance without any ...
A tiny optical booster built directly onto a chip is pointing toward a future where light, not just electricity, does the ...
We introduce ACE-Step, a novel open-source foundation model for music generation that overcomes key limitations of existing approaches and achieves state-of-the-art performance through a holistic ...
A meta-analysis pooled data on dose-response associations between daily steps and a broad range of health outcomes. Increasing daily step counts above 2,000 was associated with risk reductions in ...
Knowing the correct phase connectivity information plays a significant role in maintaining high-quality power and reliable electricity supply to end-consumers. However, managing the consumer-phase ...
In this paper, we tackle the high computational overhead of transformers for lightweight image super-resolution. (SR). Motivated by the observations of self-attention's inter-layer repetition, we ...
Sea surface temperature (SST) serves as a critical indicator for assessing marine ecosystem health. Given the increasing human exploitation of marine resources, accurate SST prediction has garnered ...