no code implementations • 29 Mar 2023 • Zhiyi Li, Douglas Orr, Valeriu Ohan, Godfrey Da Costa, Tom Murray, Adam Sanders, Deniz Beker, Dominic Masters
Furthermore, static sparsity in general outperforms dynamic sparsity.
1 code implementation • 6 Feb 2023 • Dominic Masters, Josef Dean, Kerstin Klaser, Zhiyi Li, Sam Maddrell-Mander, Adam Sanders, Hatem Helal, Deniz Beker, Andrew Fitzgibbon, Shenyang Huang, Ladislav Rampášek, Dominique Beaini
We present GPS++, a hybrid Message Passing Neural Network / Graph Transformer model for molecular property prediction.
1 code implementation • 18 Nov 2022 • Dominic Masters, Josef Dean, Kerstin Klaser, Zhiyi Li, Sam Maddrell-Mander, Adam Sanders, Hatem Helal, Deniz Beker, Ladislav Rampášek, Dominique Beaini
This technical report presents GPS++, the first-place solution to the Open Graph Benchmark Large-Scale Challenge (OGB-LSC 2022) for the PCQM4Mv2 molecular property prediction task.
no code implementations • ECCV 2020 • Deniz Beker, Hiroharu Kato, Mihai Adrian Morariu, Takahiro Ando, Toru Matsuoka, Wadim Kehl, Adrien Gaidon
3D object detection from monocular images is an ill-posed problem due to the projective entanglement of depth and scale.
3D Object Detection 3D Object Detection From Monocular Images +5
no code implementations • 22 Jun 2020 • Hiroharu Kato, Deniz Beker, Mihai Morariu, Takahiro Ando, Toru Matsuoka, Wadim Kehl, Adrien Gaidon
Deep neural networks (DNNs) have shown remarkable performance improvements on vision-related tasks such as object detection or image segmentation.