From Recognition in Brain to Recognition in Perceptual Vision Systems
Bibliographic record
Abstract
This presentation summarizes the effort of our group in designing a non Von-Neumann, biologically motivated approach to video processing and recognition. The need for such an approach is seen from the very fact that, while for humans recognition in video is easy, most recognition approaches developed to date still perform very poorly when applied to video data. Instead of focusing the effort on making the video data of better quality, as suggested by the Face Recognition Vendor Test Grand Challenge, we start from the premise that video data is inherently of bad quality, and hence the effort should be directed towards developing approaches which can deal with such low-quality data. We build our approach using the neuro-associative mechanism which is known to be of prime importance for biological vision recognition systems in enabling the accumulation of learning data in time and which is implemented by means of tuning the synaptic connections of a multiconnected neural network. As a result, we build a memorization-recognition system which can memorize a face from a video sequence and then identify this face in another video sequence. The testing of the system is deliberately done on video of very low quality, i.e. such that is just sufficient for humans to identify the faces. The application of the system for identifying computer users using a low-resolution webcam is shown. Other applications include video-annotation of TV programs and tele-conferences, and fusion of hard and soft biometrics for homeland security.
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 | 0.001 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.001 | 0.006 |
| Scholarly communication | 0.004 | 0.008 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.005 | 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".