videomae-base-finetuned-ucf101

Model Card for videomae-base-finetuned-ucf101

A WandB report here for metrics.

Table of Contents

1. Model Details
2. Uses
3. Bias, Risks, and Limitations
4. Training Details
5. Evaluation
6. Model Examination
7. Environmental Impact
8. Technical Specifications
9. Citation
10. Glossary
11. More Information
12. Model Card Authors
13. Model Card Contact
14. How To Get Started With the Model

Model Details

Model Description

<!-- Provide a longer summary of what this model is. -->

VideoMAE Base model fine tuned on UCF101

• Developed by: @nateraw
• Shared by [optional]: [More Information Needed]
• Model type: fine-tuned
• Language(s) (NLP): en
• License: mit
• Related Models [optional]: [More Information Needed]
  • Parent Model [optional]: MCG-NJU/videomae-base
• Resources for more information: [More Information Needed]

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

This model can be used for Video Action Recognition

Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

[More Information Needed]

Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

[More Information Needed]

Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

[More Information Needed]

Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recomendations.

Training Details

Training Data

<!-- This should link to a Data Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

[More Information Needed]

Training Procedure [optional]

<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->

Preprocessing

We sampled clips from the videos of 64 frames, then took a uniform sample of those frames to get 16 frame inputs for the model. During training, we used PyTorchVideo's MixVideo to apply mixup/cutmix.

Speeds, Sizes, Times

<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->

[More Information Needed]

Evaluation

<!-- This section describes the evaluation protocols and provides the results. -->

Testing Data, Factors & Metrics

Testing Data

<!-- This should link to a Data Card if possible. -->

[More Information Needed]

Factors

<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->

[More Information Needed]

Metrics

<!-- These are the evaluation metrics being used, ideally with a description of why. -->

[More Information Needed]

Results

We only trained/evaluated one fold from the UCF101 annotations. Unlike in the VideoMAE paper, we did not perform inference over multiple crops/segments of validation videos, so the results are likely slightly lower than what you would get if you did that too.

• Eval Accuracy: 0.758209764957428
• Eval Accuracy Top 5: 0.8983050584793091

Model Examination [optional]

<!-- Relevant interpretability work for the model goes here -->

[More Information Needed]

Environmental Impact

<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->

Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).

• Hardware Type: [More Information Needed]
• Hours used: [More Information Needed]
• Cloud Provider: [More Information Needed]
• Compute Region: [More Information Needed]
• Carbon Emitted: [More Information Needed]

Technical Specifications [optional]

Model Architecture and Objective

[More Information Needed]

Compute Infrastructure

[More Information Needed]

Hardware

[More Information Needed]

Software

[More Information Needed]

Citation [optional]

<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->

BibTeX:

[More Information Needed]

APA:

[More Information Needed]

Glossary [optional]

<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->

[More Information Needed]

More Information [optional]

[More Information Needed]

Model Card Authors [optional]

@nateraw

Model Card Contact

@nateraw

How to Get Started with the Model

Use the code below to get started with the model.

<details> <summary> Click to expand </summary>

from decord import VideoReader, cpu
import torch
import numpy as np

from transformers import VideoMAEFeatureExtractor, VideoMAEForVideoClassification
from huggingface_hub import hf_hub_download

np.random.seed(0)


def sample_frame_indices(clip_len, frame_sample_rate, seg_len):
    converted_len = int(clip_len * frame_sample_rate)
    end_idx = np.random.randint(converted_len, seg_len)
    start_idx = end_idx - converted_len
    indices = np.linspace(start_idx, end_idx, num=clip_len)
    indices = np.clip(indices, start_idx, end_idx - 1).astype(np.int64)
    return indices


# video clip consists of 300 frames (10 seconds at 30 FPS)
file_path = hf_hub_download(
    repo_id="nateraw/dino-clips", filename="archery.mp4", repo_type="space"
)
videoreader = VideoReader(file_path, num_threads=1, ctx=cpu(0))

# sample 16 frames
videoreader.seek(0)
indices = sample_frame_indices(clip_len=16, frame_sample_rate=4, seg_len=len(videoreader))
video = videoreader.get_batch(indices).asnumpy()

feature_extractor = VideoMAEFeatureExtractor.from_pretrained("nateraw/videomae-base-finetuned-ucf101")
model = VideoMAEForVideoClassification.from_pretrained("nateraw/videomae-base-finetuned-ucf101")

inputs = feature_extractor(list(video), return_tensors="pt")

with torch.no_grad():
    outputs = model(**inputs)
    logits = outputs.logits

# model predicts one of the 101 UCF101 classes
predicted_label = logits.argmax(-1).item()
print(model.config.id2label[predicted_label])

</details>