{"id":921,"date":"2024-12-05T18:14:08","date_gmt":"2024-12-05T18:14:08","guid":{"rendered":"https:\/\/labs.cs.queensu.ca\/istudio\/?post_type=qsc_member&p=921"},"modified":"2026-02-05T19:23:05","modified_gmt":"2026-02-05T19:23:05","slug":"victoria-armstrong","status":"publish","type":"qsc_member","link":"https:\/\/labs.cs.queensu.ca\/istudio\/members\/victoria-armstrong\/","title":{"rendered":"Victoria Armstrong\u200b\u200b\u200b"},"content":{"rendered":"
\n\t
\n\t\t\"\"\n\t<\/div>\n\t
\n\t\t

Victoria Armstrong\u200b\u200b\u200b<\/h1><\/div>\n\t\t
MSc Student<\/div>\n\t\t
School of Computing<\/div>\n\t\t
Queen’s University<\/div>\n\t\t
Member from 2020<\/em> to 2022<\/em><\/div>\n\t\t
\n\t\t\t
\n\t\t\t<\/i><\/a>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/div>\n
\n\t\n

Victoria finished her MSc at iStudio, Queen\u2019s School of Computing in 2022. She looked into designing wearables that support people with stress and anxiety physiological symptoms using AI predictive modeling. Victoria is now pursuing her PhD at Queen’s School of Computing.<\/p>\n\n\n\n

Publications<\/strong><\/h2>\n\n\n\n
<\/a><\/form>

Xiaotian Liu, Victoria Armstrong, Sara Nabil, Christian Muise<\/p>

Exploring multi-view perspectives on deep reinforcement learning agents for embodied object navigation in virtual home environments<\/a> Proceedings Article<\/span> <\/p>

In: <\/span>Proceedings of the 31st Annual International Conference on Computer Science and Software Engineering, <\/span>pp. 190\u2013195, <\/span>IBM Corp., <\/span>Toronto, Canada, <\/span>2021<\/span>.<\/p>

Abstract<\/a><\/span> | Links<\/a><\/span> | BibTeX<\/a><\/span><\/p>

@inproceedings{10.5555\/3507788.3507815,
\r\ntitle = {Exploring multi-view perspectives on deep reinforcement learning agents for embodied object navigation in virtual home environments},
\r\nauthor = {Xiaotian Liu and Victoria Armstrong and Sara Nabil and Christian Muise},
\r\nurl = {https:\/\/labs.cs.queensu.ca\/istudio\/wp-content\/uploads\/sites\/12\/2021\/08\/CASCON_x_EVOKE_Armstrong21.pdf},
\r\nyear  = {2021},
\r\ndate = {2021-01-01},
\r\nurldate = {2021-01-01},
\r\nbooktitle = {Proceedings of the 31st Annual International Conference on Computer Science and Software Engineering},
\r\npages = {190\u2013195},
\r\npublisher = {IBM Corp.},
\r\naddress = {Toronto, Canada},
\r\nseries = {CASCON '21},
\r\nabstract = {Recent years have brought the exploration of embodied reinforcement learning agents in a variety of domains. One of the advantages of artificial agents is that they can obtain visual inputs simultaneously using multiple input devices. This work explores multi-view reinforcement learning for object navigation tasks in 3D rendered virtual home environments using AI2-THOR. We trained CNN based Deep Q-learning embodied agents with egocentric, allocentric, and combined egocentric-allocentric perspectives to locate an object in an unknown environment. We compared the results of the three RL agents, and evaluated them by both reward improvement rate, and reward obtained. We demonstrate that the egocentric perspective allows for faster reward accumulation in the earlier episodes, whereas the allocentric agents obtained better long-term rewards. Interesting results arise from the combined allocentric and egocentric perspective, where we found that the agent had the best overall results by harnessing the benefits of each perspective. The results show that while single perspective embodied agents each have their own advantages, combining both inputs yields the best overall reward. Our findings provide a foundation and benchmark for building embodied RL agents with multi-view perspectives.},
\r\nkeywords = {},
\r\npubstate = {published},
\r\ntppubtype = {inproceedings}
\r\n}
\r\n<\/pre><\/div>
Close<\/a><\/p><\/div>
Recent years have brought the exploration of embodied reinforcement learning agents in a variety of domains. One of the advantages of artificial agents is that they can obtain visual inputs simultaneously using multiple input devices. This work explores multi-view reinforcement learning for object navigation tasks in 3D rendered virtual home environments using AI2-THOR. We trained CNN based Deep Q-learning embodied agents with egocentric, allocentric, and combined egocentric-allocentric perspectives to locate an object in an unknown environment. We compared the results of the three RL agents, and evaluated them by both reward improvement rate, and reward obtained. We demonstrate that the egocentric perspective allows for faster reward accumulation in the earlier episodes, whereas the allocentric agents obtained better long-term rewards. Interesting results arise from the combined allocentric and egocentric perspective, where we found that the agent had the best overall results by harnessing the benefits of each perspective. The results show that while single perspective embodied agents each have their own advantages, combining both inputs yields the best overall reward. Our findings provide a foundation and benchmark for building embodied RL agents with multi-view perspectives.<\/div>
Close<\/a><\/p><\/div>