You're looking at a specific version of this model. Jump to the model overview.
barakplasma /lingbot-map-long-video:d9e932df
Input schema
The fields you can use to run this model with an API. If you don’t give a value for a field its default value will be used.
| Field | Type | Default value | Description |
|---|---|---|---|
| video |
string
|
Video to reconstruct. Hosted URLs are recommended for large files.
|
|
| sample_fps |
number
|
5
Min: 0.25 Max: 30 |
Frames sampled per second across the full video.
|
| max_frames |
integer
|
12000
Min: 2 Max: 30000 |
Safety cap spread uniformly over the whole video.
|
| keyframe_interval |
integer
|
2
Min: 1 Max: 30 |
Persist every Nth frame in KV cache.
|
| window_size |
integer
|
128
Min: 16 Max: 256 |
KV-cache slots per upstream window.
|
| overlap_keyframes |
integer
|
8
Min: 2 Max: 32 |
Keyframes shared across windows/chunks.
|
| chunk_frames |
integer
|
512
Min: 128 Max: 1536 |
Maximum dense prediction frames held at once.
|
| point_frame_stride |
integer
|
2
Min: 1 Max: 30 |
Use every Nth frame for point geometry.
|
| spatial_stride |
integer
|
4
Min: 1 Max: 16 |
Use every Nth pixel in each dimension for export.
|
| confidence_percentile |
number
|
60
Max: 99 |
Drop this lowest-confidence percentage per frame.
|
| max_points |
integer
|
3000000
Min: 10000 Max: 8000000 |
Maximum points in the GLB.
|
| show_trajectory |
boolean
|
True
|
Include the camera path in the GLB.
|
Output schema
The shape of the response you’ll get when you run this model with an API.
{'items': {'format': 'uri', 'type': 'string'},
'title': 'Output',
'type': 'array'}