Image sentiment

I am working on a project that identify the sentiment of an image using deep learning. I have read a few papers such as the following:

  1. Visual Sentiment Prediction with Deep Convolutional Neural Networks by Xu et al.
  2. Audio-Visual Sentiment Analysis for Learning Emotional Arcs in Movies by Chu et al.
  3. From Pixels to Sentiment: Fine-tuning CNNs for Visual Sentiment Prediction by Campos et al.
    These papers are several years old. I have not seen any prominent recent papers. (My apologies, if I have missed any.) So the CNNs they have used are a bit outdated.

I am planning to use yolov8 with transfer learning to identify the sentiment. Is that approach correct? Where can I get an image sentiment dataset to train the new model? Please advise. Thank you!

As a general suggestion you can refer to dataset on kaggle.
for yolv8 you may need to train it, further you can use zero shot detectors which works without training them.