Datasets:

PSewmuthu
/

Emotion_Video_Facial_Landmarks

Tasks:

Image Classification

Modalities:

Formats:

Sub-tasks:

multi-class-image-classification

Languages:

Size:

Tags:

computer-vision

affective-computing

facial-landmarks

emotion-recognition

feature-extraction

DOI:

Libraries:

License:

Dataset card Data Studio Files Files and versions

PSewmuthu commited on Oct 19, 2025

Commit

4f2185c

·

verified ·

1 Parent(s): b6624ab

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -80,7 +80,7 @@ The features were extracted using the **MediaPipe Face Landmarker** model.
 1.  **Frame Extraction:** Each video file was processed frame-by-frame.
 2.  **Landmark Detection:** For each frame, the 478 facial landmarks were detected.
-3.  **Normalization:** All coordinates ($x, y, z$) are normalized to the range $[0.0, 1.0]$ relative to the bounding box of the face or the original frame dimensions.
 ---

 1.  **Frame Extraction:** Each video file was processed frame-by-frame.
 2.  **Landmark Detection:** For each frame, the 478 facial landmarks were detected.
+3.  **Normalization:** All coordinates (x, y, z) are normalized to the range [0.0, 1.0] relative to the bounding box of the face or the original frame dimensions.
 ---