Vadim Lebedev

Learn More
We propose a simple two-step approach for speeding up convolution layers within large convolutional neural networks based on tensor decomposition and discrim-inative fine-tuning. Given a layer, we use non-linear least squares to compute a low-rank CP-decomposition of the 4D convolution kernel tensor into a sum of a small number of rank-one tensors. At the(More)
Gatys et al. recently demonstrated that deep networks can generate beautiful textures and stylized images from a single texture example. However , their methods require a slow and memory-consuming optimization process. We propose here an alternative approach that moves the computational burden to a learning stage. Given a single example of a texture, our(More)
We propose a new approach to designing visual markers (analogous to QR-codes, markers for augmented reality, and robotic fiducial tags) based on the advances in deep generative networks. In our approach, the markers are obtained as color images synthesized by a deep network from input bit strings, whereas another deep network is trained to recover the bit(More)
Experiments based on neuronal cell-transistor couplings were made from some groups during the last years. Pioneering work in this field was carried out by Fromherz and his group (Fromherz, 2003; Schmidtner and Fromherz, 2006). We were interested of the interaction of nerve cells to serine hydrolase inhibitor diisopropylfluorophosphate (DFP), monitored by(More)
  • 1