Dependence-Aware Multitask Scheduling for Edge Video Analytics With Accuracy Guarantee
Bibliographic record
Abstract
In this paper, we investigate the optimal configuration and dependence-aware task assignment for multi-task edge video analytics. Multi-task video analytics involves multiple objects in video frames and multiple dependent tasks, resulting in existing video configuration and task assignment scheme for single-task unsuitable to this scenario. Our paper aims to efficiently assign dependent tasks to multiple collaborative edge nodes with appropriate video configuration, to achieve low latency while maintain accuracy. Firstly, we conduct extensive experiments on real-world video datasets. The results reveal that the impact of resolution on the detection accuracy varies among different sizes of objects. Moreover, the computing and communication load of dependent tasks varies along the time due to the dynamic video content. Based on the experimental results, we propose a threshold-based downsampling strategy for large objects, aiming at minimizing the transmission latency while guaranteeing task analytic accuracy. In addition, the number of objects and workload of subsequent tasks turn out to be highly correlated, the computation and transmission demands of tasks can be thus estimated for each video chunk. Then, a heuristic dependence-aware task assignment algorithm is proposed to achieve minimum completion time of dependent tasks. Experimental results demonstrate that the proposed scheme can effectively reduce the execution time of multiple tasks while guaranteeing the analytic accuracy, outperforming the state-of-the-art benchmarks.
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 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 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 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".