Who took the shot?
Team recognition and jersey reading link candidate baskets to players. You can fill in unread numbers.
SPOT Vision AI · basketball highlights & analysis
Upload your game.
Your highlights are ready.
SPOT Vision AI analyses your basketball game video: it finds every scoring play, reads the shooter’s jersey number and cuts each one into a clip. Review the results, fix anything we got wrong and download your highlight reel.
Player filters · Editable results · MP4 highlights
00:00
How it works
From camera to highlight: four acts
01 · DETECT
Find the scoring moment
Ball and rim detections identify candidate scoring moments; player tracking examines the relevant windows.
Behind the scenes: video to clip
01
Normalize
Video preprocessing
02
Detect
Object detection
03
Track
Continuous tracking
04
Identify
Team + jersey number
05
Court Map
Court mapping
06
Shots
Shot detection
07
Stats
Stats engine
08
Clips
Highlight compilation
Features
Your game’s key moments, within reach
Illustrative product previews.
Make list Makes? Review
Find scoring moments
Review candidate baskets alongside their clips, with 1, 2 or 3-point classification when court evidence supports it.
● 3PT● 2PT
▶ REELAutomatic highlights
Every basket is cut at ±5 seconds automatically; a personal reel is compiled from one player’s moments.
-5s before00:12+5s after
Clip editing
Select, trim or extend clips; compile them into a downloadable MP4 with transitions and player info cards.
#11 · Player cardDEMO
4
Makes
9
Points
4
Clips
Visible evidence
For every make you see what the decision rested on: the action box or the ball arc, whether the team check agreed, how confident the number reading was.
#112PT✓
#?3PT?
#72PT✓
Correctable results
Correct player numbers and scoring types, or hide an incorrect event. You stay in control.
Capabilities
One scoring moment, four clear steps
Not a wall of technical features. The real product flow that turns game footage into player-focused highlights.
01
Find
02
Map
03
Verify
04
Share
01
01 · Find
Catch the scoring moment
Ball and rim motion surfaces candidate baskets; player tracking focuses on the relevant event window.
Instead of scrubbing through the entire game, you start with the clips that need review.
- Timestamped scoring candidates
- Player tracking around the event
- A review flow that separates false detections
RF-DETRSAM2Event window
Real product screen02
02 · Map
From camera view to court position
The court model detects lines and landmarks. With enough confidence, the frame is mapped to FIBA court geometry.
When position is unreliable, SPOT does not invent a shot type; the event can remain unknown for review.
- 33 court landmarks
- Homography from at least 6 reliable points
- Both baskets folded onto one half-court chart
FIBAHomographyRANSAC
Real product screen03
03 · Verify
Confirm the player with your own eyes
Team appearance and jersey reading suggest a player. Watch the clip, accept the match or correct it.
Uncertainty stays visible: unread jerseys and conflicting events remain in a separate review queue.
- Player suggestion from jersey number
- Fast yes / no review
- Player and scoring-type corrections
Jersey OCRRosterHuman review
Real product screen04
04 · Share
Build a highlight from verified moments
Verified baskets are grouped by player. Select the clips you want, adjust them and assemble a single reel.
A focused game view that links totals back to video—without imaginary season percentages.
- Per-player baskets and points
- Adjustable clip in and out points
- MP4 reel from selected clips
Clip selectionReelMP4
Real product screenReal product screenImages are taken from the current SPOT product. Numbers belong to sample analyses.
By the numbers
From game footage to shareable moments
Review the moments found in your footage and build your selection without repeatedly searching the full game.
1 / 2 / 3
Scoring types
From court and event evidence
#
Player filters
By jersey number
MP4
Highlight export
A reel from selected clips
±5 sec
Event clips
See the lead-up and the finish
Behind the scenes: video to clip
- 01Video
- 02Candidates
- 03Tracking
- 04Jersey
- 05Court
- 06Review
- 07Clip
- 08Reel
SPOT vs Manual Analysis
Stop spending your weekends in video editors
Automate searching and clip collection. Make the final call by reviewing the footage.
Find a momentAutomated
Manual Analysis
Search through the video
SPOT
Timestamped scoring candidates
Choose a playerAutomated
Manual Analysis
Sort moments by hand
SPOT
Filter by jersey number
Prepare clipsAutomated
Manual Analysis
Cut each moment manually
SPOT
Automatic event clips
Build a highlightAutomated
Manual Analysis
Assemble clips in an editor
SPOT
MP4 reel from selected clips
Verify resultsYour call
Manual Analysis
Review the footage
SPOT
Review, correct or hide
Automation + your judgement — SPOT gathers the moments; you review the results and choose what to share.
Dashboard
Your game. Your call.
Watch scoring candidates, focus on a player and choose your highlight moments. Try the game-review workflow in the demo below.
SPOT/ Game review
Interactive demo · sample dataFound moments
4 sample eventsThis showcase uses sample events and is not connected to your account or games.
#67 · 2PTDemo footage
Review first. Then select.
In a real analysis, player numbers and scoring types can be corrected. Try the sample clip selection here.
1 sample clips selectedMP4 · reel workflow
Who it’s for
For everyone who is part of the game
Players
Find your moments, watch them back and create a highlight to share.
Teammates
Review game moments, check player assignments and choose your clips.
Families
Find the moments you care about in a long recording using a jersey number.
FAQ
Frequently asked questions
Straight answers about how the product works today.
It finds candidate baskets, attempts to identify the shooter and scoring type, and prepares event clips. Review and correct the results, then build a reel from selected clips.
The game ends. Your moments stay.
From your video to clips. From clips to your highlight.
Explore SPOT