Policy and Value Deep RL for Temporal Language-Agnostic Street Image Captioning
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
Dashcams and street cameras capture a huge amount of street and traffic video data. Temporal image captioning for such video data has been approached with an encoder-decoder framework and achieved substantial success in captioning accuracy. Most recently, policy and value deep reinforcement learning (PVRL) emerged as an outperformer over other decision-making frameworks for image captioning. In this paper, we design a framework that utilizes PVRL on an inhouse dataset containing temporal images of East Asia streets as a step towards designing a language-agnostic street image captioning framework that is capable of captioning temporal images of any street regardless of location. For language- invariance, the framework includes cross-modal retrieval at the character level so that similar words in different languages but in the same word-embedding space are grouped together. Our results show that PVRL can be applied successfully to temporal video captured the streets and achieve natural semantic captions; preliminary studies on our dataset suggest that the framework is capable of use in multi-language scenarios.
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