c44f6f17-7212-42d3-8b69-2556db26ae5a

The reason the script was unable to find the two videos as duplicates is that it relies on hashing the file content. Hashing calculates a unique hash value for the binary data of each file, and even small differences in the file's content (like resolution, codec, metadata, or file format) will produce a completely different hash. Consequently, videos that are the same length but have differences in quality, resolution, or encoding will not be identified as duplicates using this approach.

To identify duplicate videos that differ in resolution, format, or quality but share the same content, you'd to do more than hash-based matching. You need to employ content-based video similarity detection, which requires analyzing the actual visual or audio content of the videos. This can be achieved using specialized libraries or tools, such as:

1. Perceptual Hashing (pHash): Perceptual hashing generates a hash value based on the video's visual content, making it resilient to differences in resolution or encoding. Libraries like ImageHash (for images) or OpenCV can be adapted to compare video content.

2. Frame Comparison: You can extract key frames from the videos and compare them using image similarity techniques, such as structural similarity index (SSIM) or feature matching with OpenCV.

3. FFmpeg and Scene Detection: Extract representative frames from each video using FFmpeg and compare those frames to find similar videos.

4. Third-Party Tools: Tools like dHash or pHash libraries (or advanced APIs like PySceneDetect) are designed for this purpose.