Images and deep learning in human and urban infrastructure interactions pertinent to sustainable urban studies: Review and perspective
Bibliographic record
Abstract
• A total of 3,552 papers from 2013 to 2023 are systematically reviewed and analyzed. • We found that the cross-applications of deep learning are not standardized. • Data fusion about real-world dynamic interactions is scarce. • Four future research directions are proposed. As global urbanization intensifies, conflicts between humans and urban infrastructure increasingly affect socio-economic and environmental sustainability. Recently, using image data and deep learning to investigate the interactions between humans and urban infrastructure has been a popular approach since the fast development of Artificial Intelligence (AI). However, the convergence of data fusion, deep learning, and human-urban infrastructure interaction studies remains underexplored. Here we systematically analyze 3,552 papers from 2013 to 2023 that use image data to investigate the intersection area of data fusion, deep learning, and human and urban infrastructure interactions, aiming to elucidate the relationships among these three key elements. We found that the cross-applications of deep learning in the papers reviewed are not standardized. Given the trend of diversified data fusion, data fusion about real-world dynamic interactions is scarce. Lastly, four potential future research directions are identified: (1) understanding the dynamic and complex interaction processes; (2) exploring the potential and standards for the application of deep learning; (3) focusing more on research concerning cities in the Global South; (4) establishing suitable training datasets for the interaction between urban infrastructures and humans, which may provide valuable insights for applying foundation models in future urban studies.
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 | 0.003 | 0.011 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.008 | 0.009 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
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".