Face Attribute Prediction in Live Video using Fusion of Features and Deep Neural Networks
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Face attribute analysis from live video is a valuable aide in biometric-based person identification. This is a challenging task due to variations in lighting, occlusion, pose and other variables. To address it, we propose an effective and robust approach: extract the face features using certain selected layers of the pre-trained Convolutional Neural Network (CNN) models such as AlexNet, GoogleNet and ResNet50. We focus on the intermediate CNN layers, since the reported experimental results suggest that the best results may not always be obtained when extracting deep features using the fully connected layers. Next, we train a linear SVM on the extracted features to perform the attribute classification. We also apply a feature level fusion by concatenating the features extracted from the intermediate layers of the aforementioned networks. Our approach applied on live video achieves an average accuracy of 89.40% using the fused features which is better than the results (between 86.6% and 87%) reported for the CNNs applied only on static images.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it