no code implementations • 26 May 2024 • Zechun Liu, Changsheng Zhao, Igor Fedorov, Bilge Soran, Dhruv Choudhary, Raghuraman Krishnamoorthi, Vikas Chandra, Yuandong Tian, Tijmen Blankevoort
In this work, we identify a collection of applicable rotation parameterizations that lead to identical outputs in full-precision Transformer architectures, and find that some random rotations lead to much better quantization than others, with an up to 13 points difference in downstream zero-shot reasoning performance.
no code implementations • 11 Dec 2023 • Balakrishnan Varadarajan, Bilge Soran, Forrest Iandola, Xiaoyu Xiang, Yunyang Xiong, Lemeng Wu, Chenchen Zhu, Raghuraman Krishnamoorthi, Vikas Chandra
A common user expectation is that a click on a specific part of an object will result in the segmentation of the entire object.
no code implementations • ICCV 2015 • Bilge Soran, Ali Farhadi, Linda Shapiro
The overall prediction accuracy is 46. 2% when only 10 frames of an action are seen (2/3 of a sec).