Search Results for author: Muhammad Abdullah Jamal

Found 10 papers, 1 papers with code

M$^{3}$3D: Learning 3D priors using Multi-Modal Masked Autoencoders for 2D image and video understanding

no code implementations26 Sep 2023 Muhammad Abdullah Jamal, Omid Mohareri

We present a new pre-training strategy called M$^{3}$3D ($\underline{M}$ulti-$\underline{M}$odal $\underline{M}$asked $\underline{3D}$) built based on Multi-modal masked autoencoders that can leverage 3D priors and learned cross-modal representations in RGB-D data.

2D Semantic Segmentation Action Detection +8

SurgMAE: Masked Autoencoders for Long Surgical Video Analysis

no code implementations19 May 2023 Muhammad Abdullah Jamal, Omid Mohareri

There has been a growing interest in using deep learning models for processing long surgical videos, in order to automatically detect clinical/operational activities and extract metrics that can enable workflow efficiency tools and applications.

Self-Supervised Learning

Multi-Modal Unsupervised Pre-Training for Surgical Operating Room Workflow Analysis

no code implementations16 Jul 2022 Muhammad Abdullah Jamal, Omid Mohareri

Data-driven approaches to assist operating room (OR) workflow analysis depend on large curated datasets that are time consuming and expensive to collect.

Activity Recognition Self-Supervised Learning +2

An Empirical Study on Activity Recognition in Long Surgical Videos

no code implementations5 May 2022 Zhuohong He, Ali Mottaghi, Aidean Sharghi, Muhammad Abdullah Jamal, Omid Mohareri

In this paper, we investigate many state-of-the-art backbones and temporal models to find architectures that yield the strongest performance for surgical activity recognition.

Action Detection Activity Detection +2

Rethinking Class-Balanced Methods for Long-Tailed Visual Recognition from a Domain Adaptation Perspective

1 code implementation CVPR 2020 Muhammad Abdullah Jamal, Matthew Brown, Ming-Hsuan Yang, Liqiang Wang, Boqing Gong

Object frequency in the real world often follows a power law, leading to a mismatch between datasets with long-tailed class distributions seen by a machine learning model and our expectation of the model to perform well on all classes.

Domain Adaptation Long-tail Learning +1

Task Agnostic Meta-Learning for Few-Shot Learning

no code implementations CVPR 2019 Muhammad Abdullah Jamal, Guo-Jun Qi

However, the generalizability on new tasks of a meta-learner could be fragile when it is over-trained on existing tasks during meta-training phase.

Few-Shot Learning General Classification

Task-Agnostic Meta-Learning for Few-shot Learning

no code implementations20 May 2018 Muhammad Abdullah Jamal, Guo-Jun Qi, Mubarak Shah

Meta-learning approaches have been proposed to tackle the few-shot learning problem. Typically, a meta-learner is trained on a variety of tasks in the hopes of being generalizable to new tasks.

Classification Few-Shot Learning +1

Cannot find the paper you are looking for? You can Submit a new open access paper.