ConductVision Enterprise
ConductVision finds and counts the animals in camera-trap, pasture and pen footage, and gives field teams and herd managers a log of each animal by type, place and time.
ConductVision tracking on every frame of this clip.
Animal Detection is ConductVision's flagship proven capability, deployed in research institutions worldwide through ConductScience's equipment heritage. It draws on over ten years of building research equipment for laboratories and field studies. The system finds and tracks individual animals across frames and records their movement and behavior. The models are trained on curated scientific datasets, not internet scrapes, ensuring reliable performance across controlled environments, open pastures, and remote wildlife corridors alike.
Outcomes
Wind, moving grass and shifting shade trigger camera traps, so many frames on a card hold no animal at all.
Field staff spend hours paging past blank frames before they can count a single animal.
Pen and pasture counts depend on someone walking the rows, usually once or twice a day.
An animal that gets out or goes missing between counts can go unnoticed for most of a day.
Animals come to waterholes, troughs and feeders at uneven times, and one observer can watch only one point.
Choices about where to add water, feed or cover rest on a few hours of watching.
A count from software is hard to accept when no one can see what was counted.
Teams re-watch whole recordings to confirm single events, which takes as long as scoring them by hand.
What the model sees
ConductVision detections on this photo.
ConductVision detections on this photo.
Endpoints
Each endpoint is defined before the pilot starts, with what is counted, its unit and how the results are broken down.
| Endpoint | How it is counted | Unit | Broken down by |
|---|---|---|---|
| PrimaryDetection rate | Independent visits by each animal type, where a new visit starts after 30 minutes with no animal of that type in view. | Unit:visits per 100 camera-days | By:Camera, animal type, hour of day |
| SecondaryEmpty-frame share | Share of triggered frames in which ConductVision finds no animal. | Unit:% | By:Camera, week |
| Head count gaps | Sampled intervals in which the count in a pen view falls below the head count on record for that pen. | Unit:events per pen-day | By:Pen, hour |
| Peak count per point | Largest number of animals seen together at a water or feed point within each hour. | Unit:animals | By:Point, hour, day |
| Review time | Minutes a reviewer takes to confirm or correct the detections in 1,000 frames. | Unit:minutes per 1,000 frames | By:Reviewer, camera |
Each pilot measures ConductVision on your own footage against a sample checked by hand, before anyone relies on its numbers.
Count agreement
Difference between the ConductVision count and a hand count by a reviewer on a sample of frames.
animals per frame (mean absolute difference) · by camera, animal type
Species agreement
Share of detected animals whose type matches the identification by an expert in a sample of frames.
% · by animal type
Missed animals
Share of animals a reviewer finds in sampled frames that ConductVision did not detect.
% · by camera, day or night
How it works
Connect
Integrate cameras from laboratory overhead rigs, trail cameras, drone feeds, or barn-mounted IP cameras. The platform supports both real-time streams and batch upload of recorded footage for retrospective analysis.
Analyze
Specialized detection models identify species, estimate body condition, and classify behaviors such as feeding, resting, grooming, and social interaction. For livestock and lab animals, individual re-identification tracks subjects without physical tags.
Act
Structured observations export directly into research data formats. Alerts trigger on predefined events, a sick animal showing lethargy, a predator entering a corridor, or an animal breaching a virtual fence.
What it detects
Elephant
Each animal in view, counted per frame and per visit
Cattle
Each animal in a pen or pasture view, for head counts and occupancy
Sheep
Each animal in a flock or pen view, counted per frame
Horse
Each animal in view, counted apart from any rider
Zebra and giraffe
Each animal at water points and on open ground
Bear
Each visit to a camera-trap station, with the frame and time
Bird
Birds on rails, feeders and pen floors, counted per frame
Dog and cat
Domestic animals that enter a watched area
Features
Use cases
Researchers spend hours manually scoring rodent behavior from video. Automated tracking and behavior classification reduce scoring time by an order of magnitude while improving inter-rater consistency.
researchpharma
Conservation teams deploy trail cameras across vast areas but lack the personnel to review millions of images. AI-powered species identification filters and catalogs captures automatically.
conservationgovernment
Agricultural operations need continuous welfare monitoring across large herds. Automated behavior and gait analysis flags animals requiring veterinary attention before conditions worsen.
agricultureveterinary
Transportation departments invest millions in wildlife crossings but struggle to measure their effectiveness. Automated monitoring quantifies crossing frequency by species and time.
transportationconservationgovernment
Fish farms need to monitor feeding behavior, swimming patterns, and mortality events across multiple pens. Underwater camera feeds are processed to detect abnormal behavior and estimate biomass.
agricultureaquaculture
Food storage and processing facilities must detect rodent and bird intrusions rapidly. Animal Detection identifies pest species in real time and triggers facility management alerts.
food processinglogisticsmanufacturing
Ecologists use drone flights to survey animal populations across inaccessible terrain. The platform processes aerial video to count individuals, identify species, and map spatial distributions.
conservationresearch
Technical overview
Accuracy and speed are measured on your footage and cameras during the pilot, and reported with the conditions they were measured under.
Related
Watches water, wildlife, vessels and smoke at rivers, wetlands and shorelines.
Tracks posture, waiting and movement from pose keypoints in public spaces.
Spots and tracks small drones in the sky over sites and perimeters.
Pilot
A pilot takes one prioritized outcome, runs on your cameras, reports the endpoints above and ends with a results review and a decision.
