no code implementations • 20 Oct 2022 • Sai Raam Venkataraman, S. Balasubramanian, R. Raghunatha Sarma
Transformation-robustness is an important feature for machine learning models that perform image classification.
no code implementations • 20 Oct 2022 • Sai Raam Venkataraman, S. Balasubramanian, R. Raghunatha Sarma
This learning of such relationships allows our model to outperform both capsule and convolutional neural network baselines on transformation-robust classification tasks.
no code implementations • 15 Mar 2022 • Sai Raam Venkatraman, Rishi Rao, S. Balasubramanian, Chandra Sekhar Vorugunti, R. Raghunatha Sarma
In order to study compositional generalisation, simple reasoning and memorisation, each scene of UOUC is annotated with up to 10 novel questions.
no code implementations • 4 Oct 2020 • Sai Raam Venkatraman, Ankit Anand, S. Balasubramanian, R. Raghunatha Sarma
We present a formal grammar description of convolutional neural networks and capsule networks that shows how capsule networks can enforce such parse-tree structures, while CNNs do not.
1 code implementation • ICLR 2020 • Sai Raam Venkataraman, S. Balasubramanian, R. Raghunatha Sarma
This is done using a trainable, equivariant function defined over a grid of group-transformations.
no code implementations • 4 Aug 2019 • Sairaam Venkatraman, S. Balasubramanian, R. Raghunatha Sarma
This is done using a trainable, equivariant function defined over a grid of group-transformations.
no code implementations • 20 Aug 2018 • Sairaam Venkatraman, S. Balasubramanian, R. Raghunatha Sarma
The problem of attempting to learn the mapping between data and labels is the crux of any machine learning task.