Real-Time FPGA-Based Object Detection with Bit-Width Adaptive Quantization
Bibliographic record
Abstract
Real-time object detection on edge devices such as drones and IoT systems requires high-throughput, low-latency inference within limited hardware resources.This paper presents a hardware-aware deployment framework for the YOLOv2-Tiny object detector on FPGA, targeting low-latency and high-throughput inference in real-time edge applications. Most existing FPGA-based object detectors either skip quantization or adopt static, uniform bit-widths without considering layer-wise latency sensitivity, leading to inefficient hardware utilization and suboptimal throughput. To address this limitation, we propose a latency-guided mixed-precision quantization scheme that assigns 4-, 6-, and 8-bit representations based on each layer’s computational demand. This assignment is driven by detailed inference-time profiling of layer latency and MAC operations on the YOLOv2-Tiny network. The proposed quantization strategy is integrated into the Vitis AI toolchain and deployed on the ZCU102 board using the DPUCZDX8G accelerator. Experimental results demonstrate the effectiveness of our method: the system achieves 41 FPS at 150 MHz, outperforming prior FPGA- and GPU-based YOLOv2-Tiny implementations by up to 25% in frame rate. The Proposed maintains detection accuracy while significantly reduces model size and resource usage, offering a practical solution for embedded vision systems under strict power and latency constraints.
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 | 0.000 | 0.001 |
| 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 | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| 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".