Low accuracy with soft-nms

[tuyenqv]

Prerequisites

Please answer the following questions for yourself before submitting an issue.

• I am using the latest TensorFlow Model Garden release and TensorFlow 2.
• [x ] I am reporting the issue to the correct repository. (Model Garden official or research directory)
• [x ] I checked to make sure that this issue has not already been filed.

1. The entire URL of the file you are using

https://github.com/tensorflow/models/tree/master/research/...

2. Describe the bug

I use soft-nms with iou_threshold=0.3 and sigma =0.5, but after training i observed that its performance is not as good as hard NMS

3. Steps to reproduce

I use sdd_resnet50_fpn from object detection api with the following changes

    post_processing {
      batch_non_max_suppression {
        score_threshold: 1e-8
        iou_threshold: 0.3
        soft_nms_sigma: 0.5
        max_detections_per_class: 100
        max_total_detections: 100
      }
      score_converter: SIGMOID
    }

4. Expected behavior

According to the paper of soft-nms, its mAP should be higher than its hard nms counterpart

5. Additional context

N/A

6. System information

• OS Platform and Distribution (e.g., Linux Ubuntu 16.04): Ubuntu 18.04
• Mobile device name if the issue happens on a mobile device: N/A
• TensorFlow installed from (source or binary): binary
• TensorFlow version (use command below): 1.15.0
• Python version: 3.7.7
• Bazel version (if compiling from source): N/A
• GCC/Compiler version (if compiling from source): N/A
• CUDA/cuDNN version: 10.1
• GPU model and memory: GTX 1080Ti