UCI Nature
UCI Nature
Wildlife Pipeline
1
Select Project
2
Connect Drive
3
Confirm
Choose a Project
Select the wildlife camera project you want to work on.
UCI Campus Reserves
Wildlife Camera Photo Database
Desert Wildlife Cameras
Future project — not yet configured
Coming Soon
Connect Google Drive
Sign in with the Google account that has access to the Wildlife Camera Photo Database.
Confirm Drive Connection
We found the following Google Drive. Please confirm this is the correct project drive before continuing.
Wildlife Camera Photo Database
[email protected] · UCI Campus Reserves
UCI Nature
UCINature
Wildlife Pipeline

Dashboard

UCI Campus Reserves · Wildlife Camera Processing

Connect Google Drive
Total Images
0
All camera locations
Processed
0
↑ 44% complete
Animals Detected
0
Across all species
Pending Review
0
Needs attention
Warnings
0
Validation issues
Processing Pipeline
Completed In Progress Pending
Upload100%
-- images
Classify44%
-- / --
Review32%
--
Validate20%
--
Export0%
Not started
Last Run Summary
0%
Success
Processed--
Success rate--
Failures--
Cameras Processed
Detection Breakdown
--
Animals
Camera Locations
Loading camera locations…
Recent Activity
Loading recent activity…
--
This Session
600
Uploading Now
4
Camera Locations
Add Images
Drop images or folders here
From SD card, USB drive, or local folder
JPG · PNG · TIFF  ·  Up to 10,000 files per batch
Upload Queue
2 batches · 6.3 GB total
IMG_0001 → IMG_1200
1,200 files · Research Park · Feb 15, 2026
Research Park 4.2 GB
100%
✓ Complete
IMG_1201 → IMG_1800
600 files · Research Park · Feb 15, 2026
Research Park 2.1 GB
60%
Uploading…
600 files remaining · ~3 min left
Assign Camera Location
Which camera took these images?
Research Park
28,450 existing
Bonita Canyon 1
24,100 existing
Bonita Canyon 2
18,900 existing
Marshtrail
31,397 existing
Session Info
Date range Feb 1 – Feb 15
Camera Reconyx HyperFire 2
Format JPG · 12 MP
Upload speed 28 MB/s
Timestamp Note
Camera clocks may drift after battery changes. Use the Validate step to correct timestamp offsets.
Connected — Wildlife Camera Photo Database
[email protected] · Select a folder and prepare the Drive source
Connect Google Drive to load available folders.
Don’t see your folder? Add it to My Drive or create a shortcut, then refresh.
Paste Folder URL or ID
Paste a Google Drive folder link or raw folder ID if it doesn’t appear in the dropdown.
Optional. Helps flat folders export to one camera/location instead of one file per image.
Limits how many files are staged from the selected Drive folder.
Selected Folder
No folder selected
Manual upload mode still works without Drive.
Drive Sync
Awaiting folder sync
Sync a Drive folder to prepare the source
Choose a folder, optionally set a camera location and sync limit, then stage it into backend storage.
Idle
0%
Sync Limit All files
Backend Staging data/staging · not staged yet
Current File —
Last Sync Result
No completed Drive sync yet
Sync results will appear here after the first run.
Drive Workflow Note
Use this tab to choose and sync the Drive source into backend staging. Run Pipeline stays on the Run Model page.
Analysis Settings
MegaDetector v5.0-a SpeciesNet
? A higher threshold reduces false positives but may increase missed detections. 80% is recommended for wildlife cameras.
Low 80% High
! Selecting a larger batch size may require significant memory. Selecting 'All images' may take a very long time or cause the system to freeze.
Affects processing speed and memory usage
Pipeline Source
Local staging
Current local-only pipeline flow remains available
-- images ready to process
Processing images…
~18 min remaining
--
Processed
--
Remaining
3.2
img / sec
12
Failures
Run History
Last 4 runs
Batch Date Images Duration Status Failures
247 / 892 reviewed 645 remaining
All Pending Confirmed Flagged
312 Human Detections
Burst Groups
Queue
8 items
All images reviewed!
Every image in the queue has been confirmed or flagged. You're ready to run validation checks on this batch.
87% confidence
Human Detected
🦊
IMG_0047.JPG 1 of 8
1 / 8
✓ Confirmed
Detection Results
Animal detected Yes
Human detected —No
SpeciesNet Classification
Coyote
87% model certainty
Image Metadata
FilenameIMG_0047.JPG
CameraResearch Park
Date / TimeFeb 10, 2026 · 09:14
Burst GroupGroup 12 · 4 imgs
StatusPending
Burst Group · 4 images
Consecutive shots within 5 seconds. Apply a decision to the whole group, or expand to review individually.
Group Summary
Best confidence92%
Species (majority)Coyote
Human detected—No
Group Action
--
Total Records
--
Valid
310
Warnings
52
Errors
Deployment Date Range
Images outside this range will be flagged as warnings
Start date cannot be in the future — images must already exist
End date must be after the start date
23 images outside deployment range
Images with timestamps before Jan 15 or after May 30 — likely due to incorrect camera clock settings. Use Batch Time Correction below to adjust.
Affected images
Timestamps compared against deployment range
Filename Camera Timestamp Issue
Batch Time Correction
Apply a time offset to correct camera clock drift
Before: 2024-03-10 01:30:00
After:  2024-03-10 02:30:00
Validation Checks
Automated checks run against the processed dataset
All required columns present
Image ID, location, date & time, species, confidence, and 4 more fields
9 / 9 columns
Image IDs are unique
No duplicate file references detected
-- unique
Date / Time within deployment range
Some images fall outside Jan 15 – May 30 window
23 outside range
ML classification complete
Some images did not complete model processing
52 unprocessed
52 images — ML processing incomplete
These images have missing or invalid model output and will not be included in the export
Filename Camera Timestamp Reason
Re-run detection on these images to resolve errors
Last validated: Feb 15, 2026 · 14:32
2 validation issues found
52 unprocessed images and 23 out-of-range timestamps — these records will be excluded from the export automatically
Export Summary
Based on current validation results
--
Records to export
After all exclusions applied
75
Human detections excluded
Flagged human detections removed
--
Burst duplicates removed
Best frame per burst group kept
By Camera Location
Export Options
Choose what to include in the output file
Remove burst duplicates
Keep only the highest-confidence image from each burst — other frames in the same group will not appear in the output
−8,462 rows
Exclude human detections
Remove images where a person was detected — recommended to protect individual privacy
−75 rows
Split output by camera location
Export 4 separate files — one per camera location — instead of one combined file
4 files
Date & Camera Filter
Optional — default exports all records
Off — exporting all dates
From date cannot be in the future
To date cannot be in the future
To date must be after From date
Google Drive Not Connected
Connect your Google Drive account to export and save results.
Export to Google Drive
Output Folder
UCI Campus Reserves / Wildlife Pipeline / Exports
[email protected]
.csv
File will be saved as uci_wildlife_2024_export.csv
Preparing export… 0%
Gathering records…
Export with validation issues?
Your dataset has 2 unresolved issues. The affected records will be automatically excluded from the export — you will not lose data.
52 images — ML processing incomplete  ·  will be excluded
23 images — outside deployment date range  ·  will be excluded
Total Detections
—
Animal images
Species Found
—
Unique species
Cameras
—
Active locations
Species Distribution
Detections Over Time
Are you sure?