3 papers
cs.LG2026
Is the reconstruction loss culprit? An attempt to outperform JEPA
Alexey Potapov, Oleg Shcherbakov, Ivan Kravchenko
We evaluate JEPA-style predictive representation learning versus reconstruction-based autoencoders on a controlled "TV-series" linear dynamical system with known latent state and a…
cs.CV2024
ExtremeMETA: High-speed Lightweight Image Segmentation Model by Remodeling Multi-channel Metamaterial Imagers
Quan Liu, Brandon T. Swartz, Ivan Kravchenko +2
Deep neural networks (DNNs) have heavily relied on traditional computational units like CPUs and GPUs. However, this conventional approach brings significant computational burdens,…
cs.CV2023
Digital Modeling on Large Kernel Metamaterial Neural Network
Quan Liu, Hanyu Zheng, Brandon T. Swartz +5
Deep neural networks (DNNs) utilized recently are physically deployed with computational units (e.g., CPUs and GPUs). Such a design might lead to a heavy computational burden, sign…