Deep Vision Essay on Whale Monitoring Published in The Journal of Ocean Technology
Deep Vision's essay "Whale Watching: Automating Aerial and Surface Level Cetacean Monitoring for Improved Population Surveys" has been published in The Journal of Ocean Technology, Vol. 16, No. 3 entitled "Machine Learning and AI Applications in Ocean Monitoring and Management".
It is a multi-disciplinary problem, and one spanning sub-surface, surface, and aerial spaces. This essay discusses Deep Vision’s contribution to improved population surveys of North Atlantic right whales through the development of artificial intelligence (AI) to automatically detect, track, and geotag this endangered species using commercial off-the-shelf (COTS) electro-optical sensors. The scalability of the technology, including its resilience under all weather conditions and its application both as a surface level, mast mounted monitoring solution for ships, and as an aerial solution for uninhabited aerial vehicles and crewed surveillance aircraft is outlined.
Conservation of at-risk species is a problem that requires cooperation, collaboration, and innovation. The problem is bigger than one company, bigger than one government agency. As skills and innovations are tested, matured, and proven, the risk to species will continue to be reduced and the activities we depend on for our economic livelihood can co-exist with the many species that share the same waters.
Download the full essay:
05 March 2021
Seiche and Deep Vision Announce Collaboration to Develop Real-time Camera System for Automated Detection, Geo-location and Quantification of Marine Mammals
Deep Vision Inc. and Seiche Ltd have signed a Memorandum of Understanding (MoU) aimed at establishing a collaborative relationship in support of their respective business interests.
Deep Vision has expertise in the field of computer vision and unsupervised machine learning; in particular, Deep Vision has developed a highly advanced, unique, passive EO/IR technology that provides robust, real-time detection and tracking of marine mammals, vessels and other maritime assets on, or near, the surface of the water.
Seiche has expertise in the field of maritime environmental monitoring; in particular Seiche has developed a unique camera system, as implemented in their Mini Dual Camera, that has been specifically designed to aid the observation and monitoring of marine mammals (and other surface, and near surface objects).
Under this MoU, the two companies will advance collaborative efforts to develop a real-time camera-based system that can automatically detect, geo-locate, and quantify the presence of marine mammals and other maritime assets. The system will be resilient under all weather conditions and operable at both surface level and above, e.g. mast mounted for ships. By leveraging state-of-the-art unsupervised machine learning techniques, the system will be truly autonomous.
"We are thrilled to be working with Seiche, one of the most respected and forward-looking environmental monitoring companies on the planet. We are confident that the proposed system, once realised, will become a significant tool in protecting marine life. We are certain that such a tool is crucial for the ultimate survival of the North Atlantic right whale.” – Alan Parslow, CEO, Deep Vision.
Mark Burnett, CEO of Seiche Ltd, comments: “We are excited to be working with Deep Vision and their cutting-edge technologies, particularly their software capabilities that can rapidly assess changes in the environment and automatically tune detection algorithms to optimise performance. Our Seiche camera systems are currently providing environmental monitoring to clients around the world and this is an important next step in the provision of this technology to further protect marine life.”
Download the full Press Release.
18 December 2020
Deep Vision wraps up a second round of funding within the Atlantic Fisheries Fund framework with the successful completion of Phase 2 of the Automatic Detection of North Atlantic Right Whales (NARW) project. Technology highlights include:
- Extended to cover wide swath of cetaceans
- Extended scope to include single image detection of cetaceans - applicable to high resolution (4k, 8k) image capture surveys
- Demonstrated resilience in varying weather conditions
- Extended to include evidence of cetacean behaviour
- Higher signal-to-noise ratio (high probable cetaceans to noise)
- On-the-fly detection of probable sites of cetacean activity in the absence of visual cetacean confirmation
- System dynamically differentiates natural oceanic features from probable biologically driven features
The technology has sparked international attention. An overview of the technology is available upon request.
18 November 2020
Deep Vision's CSO was invited to present at Dstl's Invitation Only Event DARC (Defence Autonomy Research Collaboration). Deep Vision's presentation, "Passive, Adaptable Situational Awareness for Maritime Autonomy" was very well received by both Dstl and the other collaborators in attendance.
Quote from the organisers:
"Every single presenter has been hand selected so you know you're getting the creme de la creme."
02 April 2020
Deep Vision has been selected to participate in Phase 1 of the Defence and Security Accelerator (DASA)'s "Autonomy In a Challenging Environment" project. The total value of Phase 1 is approx. $3.7 million. DASA is part of the UK Ministry of Defence.
Project manager Helen Mullender said:
"The work being funded is to mature autonomous systems with the capability to operate on demand, under all conditions that may be encountered. Military operations are undertaken in all kinds of challenging environment. The inclusion of autonomous systems in these operations will demand their ability to operate effectively and efficiently regardless of the environment."
DASA delivery manager Laurence Bickerton said:
"DASA is pleased to be working with DSTL and some of the best and brightest minds in industry to provide a step change in the capability of unmanned autonomous military systems."
31 October 2019
Deep Vision's CEO presented "Automated Spotting And Monitoring Of Right Whales Using Real-Time Unsupervised Learning" at the Unmanned Systems Canada Conference in Ottawa. The presentation was once again well attended and well received..
6 September 2019
Deep Vision's CEO presented "Automatic Spotting and Monitoring of North Atlantic Right Whales From an Unmanned Aerial Platform: Using Real-Time Unsupervised Learning In-Flight." at the MAST Northern Coasts Conference in Copenhagen, Denmark. The presentation was well attended and well received. Its relation to the detection of diesel submarine snorkels and the nature of their wakes was discussed.
2 April 2019
Deep Vision receives recognition for their work on the automatic detection of North Atlantic Right Whales (NARW). At a formal announcement, held at COVE in Dartmouth, the Honourable Keith Colwell, Minister of Fisheries and Aquaculture, Province of Nova Scotia and the Honourable Jonathan Wilkinson, Minister of Fisheries, Oceans and the Canadian Coast Guard, praised Deep Vision for work performed within the Atlantic Fisheries Fund framework
Others recognised included Premium Seafoods, We’koqma’q First Nation, Ceilidh Fishermen’s Co-op, Bill & Stanley Oyster Company, Nova Scotia Fisheries Sector Council and the Guysborough County Inshore Fishermen's Association.
More details can be found in the Official News Release.