TAEHV is a Tiny AutoEncoder for Hunyuan Video (and other similar video models). TAEHV can encode and decode latents into videos more cheaply (in time & memory) than the full-size video VAEs, at the ...
Abstract: This article presents a deep autoencoder-based methodology for unsupervised anomaly detection in centrifugal pumps under limited failure data conditions, focusing on real-world applications ...
Abstract: Deep learning methods exhibit limited performance in few-shot specific emitter identification tasks due to the scarcity of labeled training samples. To address this challenge, this letter ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results