For an end-to-end guide follow this link: https://cloud.mongodb.com/ecosystem/mixpeek

​
Video

We’ll be using vuse-generic-v1 to build a collection of 1 second interval video chunks into a 768 dimension embedding collection.

You’ll need to create an Atlas vector search index of 768 dimensions for vuse-generic-v1 model embeddings. We’re calling it video_index.

​
Ingest

from mixpeek import Mixpeek
from pymongo import MongoClient

# Initialize the Mixpeek client with your API key
mixpeek = Mixpeek("YOUR_API_KEY")
client = MongoClient("YOUR_MONGODB_URI")

# Process video chunks
processed_chunks = mixpeek.tools.video.process(
    video_source="https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/Jurassic+Park+(2).mp4",
    chunk_interval=1, # 1 second intervals
    resolution=[720, 1280]
)

for index, chunk in enumerate(processed_chunks):
    print(f"Processing video chunk: {index}")

    # embed each chunk
    embed_response = mixpeek.embed.video(
        model_id="vuse-generic-v1",
        input=chunk['base64_chunk'],
        input_type="base64"
    )

    result = {
        "start_time": chunk["start_time"],
        "end_time": chunk["end_time"],
        "embedding": embed_response['embedding']
    }
    client.db.collection.insert_one(result)

​
Text Query

query = "two boys inside a car"

embed_response = mixpeek.embed.video(
    model_id="vuse-generic-v1",
    input=query,
    input_type="text"
)

aggregation = [{
    "$vectorSearch": {
        "index": "default",
        "path": "embedding",
        "queryVector": embed_response['embedding'],
        "numCandidates": 100
        "limit": 10
    }
}]

results = client.db.collection.aggregate(aggregation)
for result in results:
    print(result)

​
Video Query

# we'll use a cartoon version of jurassic park
file_url = "https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/rabbit-jurassic.mp4"

embed_response = mixpeek.embed.video(
    model_id="vuse-generic-v1",
    input=file_url,
    input_type="url"
)

aggregation = [{
    "$vectorSearch": {
        "index": "default",
        "path": "embedding",
        "queryVector": embed_response['embedding'],
        "numCandidates": 100
        "limit": 10
    }
}]

results = client.db.collection.aggregate(aggregation)
for result in results:
    print(result)

​
Image

We’ll be using clip-v1 to build a collection of image embeddings with 512 dimensions.

You’ll need to create an Atlas vector search index of 512 dimensions for clip-v1 model embeddings. We’re calling it image_index.

​
Ingest

from mixpeek import Mixpeek
from pymongo import MongoClient

# Initialize the Mixpeek client with your API key
mixpeek = Mixpeek("YOUR_API_KEY")
client = MongoClient("YOUR_MONGODB_URI")

# List of image URLs to process
image_urls = [
    "https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/image1.jpg",
    "https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/image2.jpg",
    "https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/image3.jpg",
    "https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/image4.jpg"
]

for url in image_urls:
    print(f"Processing image: {url}")

    # Embed each image
    embed_response = mixpeek.embed.image(
        model_id="openai-clip-vit-base-patch32",
        input=url,
        input_type="url"
    )

    result = {
        "image_url": url,
        "embedding": embed_response['embedding']
    }
    client.db.image_collection.insert_one(result)

​
Text Query

query = "a cat sitting on a windowsill"

embed_response = mixpeek.embed.image(
    model_id="openai-clip-vit-base-patch32",
    input=query,
    input_type="text"
)

aggregation = [{
    "$vectorSearch": {
        "index": "image_index",
        "path": "embedding",
        "queryVector": embed_response['embedding'],
        "numCandidates": 100,
        "limit": 5
    }
}]

results = client.db.image_collection.aggregate(aggregation)
for result in results:
    print(result)

​
Image Query

# Use an image from the same bucket as a query
query_image = "https://mixpeek-public-demo.s3.us-east-2.amazonaws.com/mongodb/query_image.jpg"

embed_response = mixpeek.embed.image(
    model_id="clip-v1",
    input=query_image,
    input_type="url"
)

aggregation = [{
    "$vectorSearch": {
        "index": "image_index",
        "path": "embedding",
        "queryVector": embed_response['embedding'],
        "numCandidates": 100,
        "limit": 5
    }
}]

results = client.db.image_collection.aggregate(aggregation)
for result in results:
    print(result)

Was this page helpful?

TimescaleDBSupabase