Department of Fisheries Oceanography / SMAST seminar - November 15, 2017 - Ben Woodward
When: Wednesday,
November 15, 2017
3:30 PM
-
4:30 PM
Where: New Bedford New Bedford, MA
Description: Department of Fisheries Oceanography
Applied Deep Learning and Computer Vision for Fisheries Science
Ben Woodward
C-Vision
Wednesday, November 15, 2017
3:30 pm - 4:30 pm
SMAST E, Room 101/102
836 South Rodney French Boulevard, New Bedford, MA
Abstract: Deep Learning and Computer Vision for video and image processing has made immense progress in the last 5 years, especially in the areas of object detection and classification. The advent of modern deep learning training architectures such as Tensorflow, Caffe, and PyTorch, combined with open source applications and pre-trained models, the barrier to utilizing these powerful technologies in diverse fields is lower than ever. Concurrently with the development of these powerful tools, the cost to generate massive amounts of video and imagery data has been drastically reduced as well. As a result, many fields within the sciences are now flooded with data that there is no hope of manually reviewing for extraction of all the valuable information contained therein. In this talk I will discuss the application of Deep Learning technologies to fisheries science, and highlight two real world applications in underwater trawl surveys and electronic monitoring of discards.
To access the live broadcasting you will have to login as smast@umassd.edu with the password: smastumassd
The link to the login is:
https://echo360.org/home
Don't forget to click on I am not a robot.
After login you will have to click on ALL CLASSES (MAR 700 - 01 - DEOS Seminar or MAR 700 - 02 - DFO Seminar) and click on the green LIVE streaming.
To view a video of an SMAST seminar (post-October 1, 2014), go to http://www.umassd.edu/smast/newsandevents/seminarseries/ and click on a highlighted title.
For more information, please contact cfox@umassd.edu
Applied Deep Learning and Computer Vision for Fisheries Science
Ben Woodward
C-Vision
Wednesday, November 15, 2017
3:30 pm - 4:30 pm
SMAST E, Room 101/102
836 South Rodney French Boulevard, New Bedford, MA
Abstract: Deep Learning and Computer Vision for video and image processing has made immense progress in the last 5 years, especially in the areas of object detection and classification. The advent of modern deep learning training architectures such as Tensorflow, Caffe, and PyTorch, combined with open source applications and pre-trained models, the barrier to utilizing these powerful technologies in diverse fields is lower than ever. Concurrently with the development of these powerful tools, the cost to generate massive amounts of video and imagery data has been drastically reduced as well. As a result, many fields within the sciences are now flooded with data that there is no hope of manually reviewing for extraction of all the valuable information contained therein. In this talk I will discuss the application of Deep Learning technologies to fisheries science, and highlight two real world applications in underwater trawl surveys and electronic monitoring of discards.
To access the live broadcasting you will have to login as smast@umassd.edu with the password: smastumassd
The link to the login is:
https://echo360.org/home
Don't forget to click on I am not a robot.
After login you will have to click on ALL CLASSES (MAR 700 - 01 - DEOS Seminar or MAR 700 - 02 - DFO Seminar) and click on the green LIVE streaming.
To view a video of an SMAST seminar (post-October 1, 2014), go to http://www.umassd.edu/smast/newsandevents/seminarseries/ and click on a highlighted title.
For more information, please contact cfox@umassd.edu
Contact: > See Description for contact information
Topical Areas: School for Marine Sciences and Technology, SMAST Seminar Series