Effects of Different Pre-Training Algorithms on Video Classification Results
Bibliographic record
Abstract
The dataset used in the study consists of 600 videos obtained by us from social media (Instagram).The data set includes 6 different video files.These files contain videos of different subjects and such as dance, driving car, fitness, make-up, mukbang.Dataset, containing 600 videos, 100 videos in each class.Additionally, the length of each video is approximately 1000 and the number of frames varies accordingly.Additionally, the videos are in mp4 format and received and processed in this format.All frames were used in classification.Because video combines multiple picture frames to create a series of images, classification is performed on this image data.The image feature information of these frames was extracted with the help of three different pre-training algorithms.These features include all features in the framework.These algorithms; ResNet50, ResNet101, GoogleNet.It is possible to collect data with different algorithms during the pre-training phase, observe the results clearly with different algorithms, and classify them in MATLAB by taking 1000 features from each algorithm.To check whether the features obtained from each pre-training algorithm were classifiable, before proceeding to the classification stage, the videos were found to be classifiable by t-SNE and pixel analysis.After analyzing the features of three different pre-training algorithms, classification was started.In the classification phase, the Bi-LSTM classifier was used because it is a time-dependent function and classifies all the data after reviewing it in detail.In conclusion; The effects of videos belonging to these six different classes on the classification accuracy of the results obtained with the same parameters but different pre-training algorithms were investigated.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.026 |
| Meta-epidemiology (narrow) | 0.003 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".