Reverse Video Search
Search your video library by submitting a video clip as the query. The pipeline decomposes the query video into scene-level embeddings and matches them against your indexed video collection to find visually and semantically similar content.
from mixpeek import Mixpeekclient = Mixpeek(api_key="YOUR_API_KEY")# Create a video collection with scene decompositioncollection = client.collections.create(namespace_id="ns_your_namespace",name="video_library",extractors=["multimodal-extractor"],params={"video_chunking": "scene-based"})# Upload video catalogclient.buckets.upload(bucket_id="bkt_your_bucket",url="s3://your-bucket/videos/")# Reverse video search: query with a video clipresults = client.retrievers.execute(retriever_id="ret_your_retriever",query={"video_url": "https://example.com/query-clip.mp4"})for doc in results["results"]:print(f"Video: {doc['root_object_id']} at {doc['start_time']}s (score: {doc['score']:.3f})")
Feature Extractors
Retriever Stages
aggregate
Compute aggregations (COUNT, SUM, AVG, etc.) on pipeline results
rerank
Rerank documents using cross-encoder models for accurate relevance
