Video Frame Interpolation
FILM synthesizes smooth, high-quality intermediate frames for fluid motion in videos with significant movement.
API
If you're looking for an API, you can choose from your desired programming language.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
import requests
import base64
# Use this function to convert an image file from the filesystem to base64
def image_file_to_base64(image_path):
with open(image_path, 'rb') as f:
image_data = f.read()
return base64.b64encode(image_data).decode('utf-8')
# Use this function to fetch an image from a URL and convert it to base64
def image_url_to_base64(image_url):
response = requests.get(image_url)
image_data = response.content
return base64.b64encode(image_data).decode('utf-8')
# Use this function to convert a list of image URLs to base64
def image_urls_to_base64(image_urls):
return [image_url_to_base64(url) for url in image_urls]
api_key = "YOUR_API_KEY"
url = "https://api.segmind.com/v1/video-frame-interpolation"
# Request payload
data = {
"input_video": "https://segmind-resources.s3.amazonaws.com/output/568d37fa-1691-4ea5-aeb7-a904310a84df-2d5e91ed-66d2-4880-aa66-6aee09ad56f6.mp4",
"frame_multiplier": 2,
"output_fps": 30,
"preserve_audio": True,
"base64": False
}
headers = {'x-api-key': api_key}
response = requests.post(url, json=data, headers=headers)
print(response.content) # The response is the generated imageAttributes
Input video URL for processing
Frame multiplier for video processing
min : 1,
max : 10
Output frames per second for the processed video
min : 1,
max : 120
Whether to preserve audio from the input video
Output as base64 encoded string
To keep track of your credit usage, you can inspect the response headers of each API call. The x-remaining-credits property will indicate the number of remaining credits in your account. Ensure you monitor this value to avoid any disruptions in your API usage.
Resources to get you started
Everything you need to know to get the most out of Video Frame Interpolation
FILM: Frame Interpolation for Large Motion
FILM is an advanced AI model by Google Research designed to generate smooth, high-quality intermediate frames between widely spaced images. It leverages a scale-agnostic feature pyramid and bi-directional motion estimation to handle large movements, inpaint occlusions, and maintain temporal consistency. Use FILM to turn ordinary footage into cinematic slow motion, smooth time-lapse sequences, or crisp gaming and sports replays.
Getting Started
- •Provide a short input video URL for faster turnaround.
- •Choose your Frame Multiplier (1–10) for desired smoothness.
- •Set Output FPS (1–120) to match your target viewing experience.
- •Toggle Preserve Audio if original sound is needed.
- •(Advanced) Enable Base64 Output to embed results directly in web pages.
Parameter Recommendations
| Use Case | Frame Multiplier | Output FPS | Preserve Audio |
|---|---|---|---|
| Natural slow motion (film style) | 2–3 | 24 | ✓ |
| Sports / gaming highlights | 3–4 | 60 | Optional |
| Dramatic ultra-smooth effect | 5–8 | 60–120 | ✗ |
| Time-lapse smoothing | 2–3 | 30–60 | ✗ |
| Security footage clarity | 1–2 | 30 | Optional |
| Animation & VFX transitions | 4–6 | 24–60 | ✗ |
Examples
Cinematic slow motion (film look):
{
"input_video": "https://…/clip.mp4",
"frame_multiplier": 2,
"output_fps": 24,
"preserve_audio": true
}
Ultra-smooth gaming replay:
{
"input_video": "https://…/game.mp4",
"frame_multiplier": 4,
"output_fps": 60,
"preserve_audio": false
}
Best Practices
- •Keep clips under 30 seconds for faster processing. Split longer videos into segments.
- •For scenes with rapid motion or complex backgrounds, use lower multipliers (2–3×) to reduce artifacts.
- •If you need maximum fluidity (e.g., slow-motion showcases), push multipliers to 6–8× but review results for occasional ghosting.
- •Preserve audio when timing and original dialogue or effects are critical; disable it when you plan to re-score or overlay new soundtracks.
- •Use Base64 output sparingly—this is ideal for quick embedding but increases payload size.
Troubleshooting
- •Soft edges or ghosting in fast pans? Lower the multiplier by 1 and reprocess.
- •Inconsistent inpainting on occluded regions? Try a two-pass approach: first interpolate at 2×, then run a higher multiplier on sections of interest.
- •Audio desync? Ensure “preserve_audio” is enabled and match the final FPS to your original timeline.
With these settings and tips, you can harness FILM’s power to produce buttery-smooth, visually stunning footage across a variety of projects.
Other Popular Models
Discover other models you might be interested in.
storydiffusion
Story Diffusion turns your written narratives into stunning image sequences.
faceswap-v2
Take a picture/gif and replace the face in it with a face of your choice. You only need one image of the desired face. No dataset, no training
sdxl1.0-txt2img
The SDXL model is the official upgrade to the v1.5 model. The model is released as open-source software
sd2.1-faceswapper
Take a picture/gif and replace the face in it with a face of your choice. You only need one image of the desired face. No dataset, no training
