1 code implementation • 20 Nov 2023 • Pooya Fayyazsanavi, Negar Nejatishahidin, Jana Kosecka
We also propose a novel two-stage inference approach that re-ranks the hypotheses using the language model capabilities of the decoder.
no code implementations • 17 Nov 2023 • Yimeng Li, Navid Rajabi, Sulabh Shrestha, Md Alimoor Reza, Jana Kosecka
We aim to develop a cost-effective labeling approach to obtain pseudo-labels for semantic segmentation and object instance detection in indoor environments, with the ultimate goal of facilitating the training of lightweight models for various downstream tasks.
no code implementations • 18 Aug 2023 • Navid Rajabi, Jana Kosecka
In this work, we show qualitatively (using explainability tools) and quantitatively (using object detectors) that the poor object localization "grounding" ability of the models is a contributing factor to the poor image-text matching performance.
no code implementations • 26 Apr 2023 • Negar Nejatishahidin, Will Hutchcroft, Manjunath Narayana, Ivaylo Boyadzhiev, Yuguang Li, Naji Khosravan, Jana Kosecka, Sing Bing Kang
In this paper, we address the problem of wide-baseline camera pose estimation from a group of 360$^\circ$ panoramas under upright-camera assumption.
no code implementations • 17 Apr 2023 • Pooya Fayyazsanavi, Zhiqiang Wan, Will Hutchcroft, Ivaylo Boyadzhiev, Yuguang Li, Jana Kosecka, Sing Bing Kang
While the existing deep learning-based room layout estimation techniques demonstrate good overall accuracy, they are less effective for distant floor-wall boundary.
1 code implementation • 17 Dec 2022 • Yimeng Li, Arnab Debnath, Gregory J. Stein, Jana Kosecka
In this work, we compare the state-of-the-art Deep Reinforcement Learning based approaches with Partially Observable Markov Decision Process (POMDP) formulation of the point goal navigation problem.
no code implementations • 15 Nov 2022 • Yimeng Li, Arnab Debnath, Gregory Stein, Jana Kosecka
Our approach surpasses the greedy strategies by 2. 1% and the RL-based exploration methods by 8. 4% in terms of coverage.
no code implementations • 4 Oct 2022 • Sulabh Shrestha, Yimeng Li, Jana Kosecka
Given the spatial and temporal consistency cues used for pixel level data association, we use a variant of contrastive learning to train a DCNN model for predicting semantic segmentation from RGB views in the target environment.
1 code implementation • 2 Mar 2022 • Negar Nejatishahidin, Pooya Fayyazsanavi, Jana Kosecka
The deep convolutional network models (CNN) for pose estimation are typically trained and evaluated on datasets specifically curated for object detection, pose estimation, or 3D reconstruction, which requires large amounts of training data.
Ranked #1 on Pose Estimation on Pix3D
no code implementations • 25 Nov 2021 • Yimeng Li, Jana Kosecka
Recent efforts in deploying Deep Neural Networks for object detection in real world applications, such as autonomous driving, assume that all relevant object classes have been observed during training.
no code implementations • 26 Jun 2020 • Ali Mirzaeian, Jana Kosecka, Houman Homayoun, Tinoosh Mohsenin, Avesta Sasan
This paper proposes an ensemble learning model that is resistant to adversarial attacks.
no code implementations • 9 Mar 2020 • Dom Huh, Sai Gurrapu, Frederick Olson, Huzefa Rangwala, Parth Pathak, Jana Kosecka
With advancements in deep model architectures, tasks in computer vision can reach optimal convergence provided proper data preprocessing and model parameter initialization.
no code implementations • ECCV 2020 • Georgios Georgakis, Ren Li, Srikrishna Karanam, Terrence Chen, Jana Kosecka, Ziyan Wu
In this work, we address this gap by proposing a new technique for regression of human parametric model that is explicitly informed by the known hierarchical structure, including joint interdependencies of the model.
1 code implementation • 4 Mar 2020 • Yimeng Li, Jana Kosecka
The advances in deep reinforcement learning recently revived interest in data-driven learning based approaches to navigation.
no code implementations • 4 Mar 2020 • Al Amin Hosain, Panneer Selvam Santhalingam, Parth Pathak, Huzefa Rangwala, Jana Kosecka
American Sign Language recognition is a difficult gesture recognition problem, characterized by fast, highly articulate gestures.
2 code implementations • 18 Nov 2019 • Georgios Georgakis, Yimeng Li, Jana Kosecka
This work presents a modular architecture for simultaneous mapping and target driven navigation in indoors environments.
no code implementations • 24 Sep 2019 • Al Amin Hosain, Panneer Selvam Santhalingam, Parth Pathak, Jana Kosecka, Huzefa Rangwala
Despite having similarity with the well-studied human activity recognition, the use of 3D skeleton data in sign language recognition is rare.
1 code implementation • ICCV 2019 • Georgios Georgakis, Srikrishna Karanam, Ziyan Wu, Jana Kosecka
In this paper, we solve this key problem of existing methods requiring expensive 3D pose annotations by proposing a new method that matches RGB images to CAD models for object pose estimation.
9 code implementations • 18 Jul 2018 • Peter Anderson, Angel Chang, Devendra Singh Chaplot, Alexey Dosovitskiy, Saurabh Gupta, Vladlen Koltun, Jana Kosecka, Jitendra Malik, Roozbeh Mottaghi, Manolis Savva, Amir R. Zamir
Skillful mobile operation in three-dimensional environments is a primary topic of study in Artificial Intelligence.
no code implementations • 8 Jun 2018 • Etienne Pot, Alexander Toshev, Jana Kosecka
In robotic applications, we often face the challenge of discovering new objects while having very little or no labelled training data.
3 code implementations • 15 May 2018 • Arsalan Mousavian, Alexander Toshev, Marek Fiser, Jana Kosecka, Ayzaan Wahid, James Davidson
We propose to using high level semantic and contextual features including segmentation and detection masks obtained by off-the-shelf state-of-the-art vision as observations and use deep network to learn the navigation policy.
1 code implementation • 13 Mar 2018 • Phil Ammirato, Cheng-Yang Fu, Mykhailo Shvets, Jana Kosecka, Alexander C. Berg
While state-of-the-art general object detectors are getting better and better, there are not many systems specifically designed to take advantage of the instance detection problem.
no code implementations • CVPR 2018 • Georgios Georgakis, Srikrishna Karanam, Ziyan Wu, Jan Ernst, Jana Kosecka
Finding correspondences between images or 3D scans is at the heart of many computer vision and image retrieval applications and is often enabled by matching local keypoint descriptors.
no code implementations • 1 Aug 2017 • Phi-Hung Le, Jana Kosecka
The paper exploits weak Manhattan constraints to parse the structure of indoor environments from RGB-D video sequences in an online setting.
no code implementations • 27 Feb 2017 • Phil Ammirato, Patrick Poirson, Eunbyung Park, Jana Kosecka, Alexander C. Berg
We present a new public dataset with a focus on simulating robotic vision tasks in everyday indoor environments using real imagery.
no code implementations • 25 Feb 2017 • Georgios Georgakis, Arsalan Mousavian, Alexander C. Berg, Jana Kosecka
In this work we explore the ability of using synthetically generated composite images for training state-of-the-art object detectors, especially for object instance detection.
11 code implementations • CVPR 2017 • Arsalan Mousavian, Dragomir Anguelov, John Flynn, Jana Kosecka
In contrast to current techniques that only regress the 3D orientation of an object, our method first regresses relatively stable 3D object properties using a deep convolutional neural network and then combines these estimates with geometric constraints provided by a 2D object bounding box to produce a complete 3D bounding box.
Ranked #9 on Vehicle Pose Estimation on KITTI Cars Hard
no code implementations • 26 Sep 2016 • Georgios Georgakis, Md. Alimoor Reza, Arsalan Mousavian, Phi-Hung Le, Jana Kosecka
This paper presents a new multi-view RGB-D dataset of nine kitchen scenes, each containing several objects in realistic cluttered environments including a subset of objects from the BigBird dataset.
no code implementations • 19 Sep 2016 • Patrick Poirson, Phil Ammirato, Cheng-Yang Fu, Wei Liu, Jana Kosecka, Alexander C. Berg
For applications in navigation and robotics, estimating the 3D pose of objects is as important as detection.
no code implementations • 1 Sep 2016 • Arsalan Mousavian, Jana Kosecka
In this work we present an approach for geo-locating a novel view and determining camera location and orientation using a map and a sparse set of geo-tagged reference views.
no code implementations • 3 Jun 2016 • Md. Alimoor Reza, Jana Kosecka
Future advancements in robot autonomy and sophistication of robotics tasks rest on robust, efficient, and task-dependent semantic understanding of the environment.
no code implementations • 25 Apr 2016 • Arsalan Mousavian, Hamed Pirsiavash, Jana Kosecka
The proposed model is trained and evaluated on NYUDepth V2 dataset outperforming the state of the art methods on semantic segmentation and achieving comparable results on the task of depth estimation.
no code implementations • 20 Sep 2015 • Arsalan Mousavian, Jana Kosecka
Several recent approaches showed how the representations learned by Convolutional Neural Networks can be repurposed for novel tasks.
no code implementations • CVPR 2013 • Gautam Singh, Jana Kosecka
This paper presents a nonparametric approach to semantic parsing using small patches and simple gradient, color and location features.