WADE: A Reasoning-Annotated Benchmark for Multi-Instance Floating-Waste Grounding with Compact Vision-Language Models
2026-08-24 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors created a new dataset called WADE to help track floating trash in rivers, focusing on pictures from rural Bangladesh. This dataset includes many labeled examples and detailed rules to help models recognize different types of waste. They tested several vision-language models to see how well these models could find, count, and explain floating trash without much training. By fine-tuning one model with their data, the authors improved its accuracy, but most trash items still went undetected, showing that this task remains difficult. Overall, the study provides a useful, challenging benchmark for future work on monitoring aquatic waste with compact AI models.
floating wasteinland waterwaysvision-language modelsdatasetbounding boxesfine-tuningzero-shot learningobject detectioncountinghallucination
Authors
Md. Asaduzzaman Shuvo, Ahsan Farabi, Md. Abdul Ahad Minhaz, Mahedi Hasan, Israt Khandaker, Ibrahim Khalil Shanto, Muhammad Nomani Kabir
Abstract
Floating waste in inland waterways threatens aquatic ecosystems and requires timely monitoring under cluttered, multi-object conditions. Existing aquatic-waste datasets provide limited geographic coverage, sparse multi-instance annotations, and little supervision beyond boxes and labels. Compact vision-language models (VLMs) therefore remain insufficiently evaluated for jointly localizing, classifying, counting, and explaining floating waste. We introduce WADE, a reasoning-annotated benchmark containing 2,167 images from rural Bangladesh, 13,608 bounding boxes, and ten waste categories. Each annotation is associated with class-level recognition rules covering visual cues, likely confusions, and discriminative features. We evaluate six VLMs under zero-shot, two-shot, reasoning-guided, and fine-tuned settings using detection, counting, and hallucination metrics. For resource-efficient adaptation, we jointly fine-tune Qwen3-VL-2B on boxes, labels, and reasoning chains using QLoRA. Fine-tuning increases recall from 0.0248 to 0.2339 and F1 from 0.0257 to 0.2163, while reducing image-level hallucination from 0.6836 to 0.0883. However, over three-quarters of instances remain undetected, establishing WADE as a challenging benchmark for dense floating-waste grounding with compact VLMs.