TY - JOUR
T1 - MAVNet
T2 - An effective semantic segmentation micro-network for MAV-based tasks
AU - Nguyen, Ty
AU - Shivakumar, Shreyas S.
AU - Miller, Ian D.
AU - Keller, James
AU - Lee, Elijah S.
AU - Zhou, Alex
AU - Özaslan, Tolga
AU - Loianno, Giuseppe
AU - Harwood, Joseph H.
AU - Wozencraft, Jennifer
AU - Taylor, Camillo J.
AU - Kumar, Vijay
N1 - Funding Information:
Manuscript received February 25, 2019; accepted June 26, 2019. Date of publication July 15, 2019; date of current version August 2, 2019. This letter was recommended for publication by Associate Editor I. Sa and Editor J. Roberts upon evaluation of the reviewers’ comments. This work was supported in part by the MAST Collaborative Technology Alliance - Contract W911NF-08-2-0004, in part by ARL Grant W911NF-08-2-0004, in part by ONR Grants N00014-07-1-0829 and N00014-14-1-0510, in part by ARO Grant W911NF-13-1-0350, in part by NSF Grants IIS-1426840 and IIS-1138847, in part by DARPA Grants HR001151626 and HR0011516850, and in part by the Semiconductor Research Corporation (SRC) and DARPA. (Corresponding author: Ty Nguyen.) T. Nguyen, S. S. Shivakumar, I. D. Miller, J. Keller, E. S. Lee, A. Zhou, T. Özaslan, C. J. Taylor, and V. Kumar are with the GRASP Lab, University of Pennsylvania, Philadelphia, PA 19104 USA (e-mail: tynguyen@seas.upenn.edu; sshreyas@seas.upenn.edu; iandm@seas.upenn.edu; jfkeller@seas.upenn.edu; elslee@seas.upenn.edu; alexzhou@seas.upenn.edu; ozaslan@seas.upenn.edu; cjtaylor@seas.upenn.edu; kumar@seas.upenn.edu).
Publisher Copyright:
© 2020 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
PY - 2019/10
Y1 - 2019/10
N2 - Real-time semantic image segmentation on platforms subject to size, weight, and power constraints is a key area of interest for air surveillance and inspection. In this letter, we propose MAVNet: a small, light-weight, deep neural network for real-time semantic segmentation on micro aerial vehicles (MAVs). MAVNet, inspired by ERFNet [E. Romera, J. M. lvarez, L. M. Bergasa, and R. Arroyo, "ErfNet: Efficient residual factorized convnet for real-time semantic segmentation," IEEE Trans. Intell. Transp. Syst., vol. 19, no. 1, pp. 263-272, Jan. 2018.], features 400 times fewer parameters and achieves comparable performancewith somereference models in empirical experiments. Additionally,we provide two novel datasets that represent challenges in semantic segmentation for real-timeMAVtracking and infrastructure inspection tasks and verify MAVNet on these datasets. Our algorithm and datasets are made publicly available.
AB - Real-time semantic image segmentation on platforms subject to size, weight, and power constraints is a key area of interest for air surveillance and inspection. In this letter, we propose MAVNet: a small, light-weight, deep neural network for real-time semantic segmentation on micro aerial vehicles (MAVs). MAVNet, inspired by ERFNet [E. Romera, J. M. lvarez, L. M. Bergasa, and R. Arroyo, "ErfNet: Efficient residual factorized convnet for real-time semantic segmentation," IEEE Trans. Intell. Transp. Syst., vol. 19, no. 1, pp. 263-272, Jan. 2018.], features 400 times fewer parameters and achieves comparable performancewith somereference models in empirical experiments. Additionally,we provide two novel datasets that represent challenges in semantic segmentation for real-timeMAVtracking and infrastructure inspection tasks and verify MAVNet on these datasets. Our algorithm and datasets are made publicly available.
KW - Aerial systems: perception and autonomy
KW - Object detection
KW - Recognition
KW - Segmentation and categorization
KW - Semantic scene understanding
KW - Semantic segmentation
UR - http://www.scopus.com/inward/record.url?scp=85078542806&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85078542806&partnerID=8YFLogxK
U2 - 10.1109/LRA.2019.2928734
DO - 10.1109/LRA.2019.2928734
M3 - Article
AN - SCOPUS:85078542806
SN - 2377-3766
VL - 4
SP - 3908
EP - 3915
JO - IEEE Robotics and Automation Letters
JF - IEEE Robotics and Automation Letters
IS - 4
M1 - 8764006
ER -