Downloads · 30 days
11
1% of all-time downloads
tiktok-video-safety-classifier
KhoiBui/tiktok-video-safety-classifier
tiktok-video-safety-classifier is a video classification model from KhoiBui. Use it for the video classification task on the model card, and read the license before you ship it in a product. The card lists the license as mit.
This model classifies TikTok video frames as safe or harmful.
All-time downloads
1.3K
Public
Parameters
86.2M
345 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors345 MB · 100%
At a glance
- Task
- Video Classification
- License
- mit
- Model type
- videomae
- Access
- Public
- Created
- Jan 28, 2026
- Updated
- Jan 28, 2026
- SHA
- 9842eb95
Datasets
Model card
From the Hugging Face model README
TikTok Content Safety - Video Classifier
This model classifies TikTok video frames as safe or harmful.
Usage
from transformers import AutoImageProcessor, VideoMAEForVideoClassification
from decord import VideoReader, cpu
import numpy as np
processor = AutoImageProcessor.from_pretrained("KhoiBui/tiktok-video-safety-classifier")
model = VideoMAEForVideoClassification.from_pretrained("KhoiBui/tiktok-video-safety-classifier")
# Load video and sample 16 frames
vr = VideoReader("video.mp4", ctx=cpu(0))
indices = np.linspace(0, len(vr) - 1, 16).astype(int)
frames = list(vr.get_batch(indices).asnumpy())
inputs = processor(frames, return_tensors="pt")
outputs = model(**inputs)
prediction = outputs.logits.argmax(-1).item() # 0=safe, 1=harmful
Model Details
- Base Model: VideoMAE (MCG-NJU/videomae-base-finetuned-kinetics)
- Task: Binary classification (safe/harmful)
- Input: 16 frames, 224x224
- Task
- Video Classification
- Type
- videomae
- License
- mit
- Languages
- vi, en
- Created
- Jan 28, 2026
- Updated
- Jan 28, 2026