Notebooks
S
Snowflake
Image Processing Pipeline

Image Processing Pipeline

data-sciencenotebookmachine-learningsnowflake-demo-notebooksdata-engineeringPythonsqlImage_Processing_Pipeline_Stream_Task_Cortex_Complete

Image Processing Pipeline using Snowflake Cortex

This notebooks demonstrates the implementation of an image processing pipeline using Streams, Tasks and SNOWFLAKE.CORTEX.COMPLETE multimodal capability. (Currently in Public Preview.)

Step 1: Create Snowflake managed stage to store sample images.

[ ]

Step 2: Download two sample images provided below and upload them on stage GENAI_IMAGES. Learn how

Sample images:

Note: Sample images provided courtesy of Dash.

Step 3: Create Stream images_stream on stage GENAI_IMAGES to detect changes.

[ ]

Step 4: Create target table image_analysis to store image analysis.

[ ]

Step 5: Preview image analysis produced on the sample images

[ ]

Step 6: Create Task image_analysis_task to process new images uploaded on stage GENAI_IMAGES using SNOWFLAKE.CORTEX.COMPLETE() multimodal capability.

[ ]

Step 7: Confirm Task status

[ ]

Step 8: Download new sample image provided below and upload it on stage GENAI_IMAGES. Learn how

Sample image:

Note: Sample image provided courtesy of Dash.

Step 9: Preview image analysis produced on the new sample image

[ ]

Step 10: Suspend task

[ ]