Search results

Metrics reproduction yields significant differences for CollisionRate

Status: Open.

#131 In swc-17/SparseDrive;

· w-poh opened on Jul 2, 2026

fix_scale 初始化bug

Status: Open.

#130 In swc-17/SparseDrive;

· xwqwenqi opened on May 20, 2026

Bug in Check Collision ?

Status: Open.

#129 In swc-17/SparseDrive;

· XZLeo opened on Apr 23, 2026

When reproducing ResNet101 version, map_normal become very low

Status: Open.

#128 In swc-17/SparseDrive;

· creatorHJJ opened on Mar 12, 2026

车道线检测和标注问题

Status: Open.

#127 In swc-17/SparseDrive;

· oym1994 opened on Jan 12, 2026

The evaluation results in Stage 2 show a large discrepancy (the learning rate has been modified).

Status: Open.

#126 In swc-17/SparseDrive;

· 13596673883 opened on Jan 9, 2026

结果复现与论文结果相差较大（The reproduced results differ significantly from those in the paper.）

Status: Open.

#124 In swc-17/SparseDrive;

· Jule-ggg opened on Nov 26, 2025

what if setting use_deformable_func to be false?

Status: Open.

#123 In swc-17/SparseDrive;

· oym1994 opened on Nov 25, 2025

Any bug for UniAD evaluation？

Status: Open.

#122 In swc-17/SparseDrive;

· zhoujiawei3 opened on Nov 21, 2025

加载权重后进行test，loss从9增加到15

Status: Open.

#121 In swc-17/SparseDrive;

· oym1994 opened on Nov 18, 2025

why head.det_head.instance_bank.instance_feature not requires grad while head.map_head.instance_bank.instance_feature requires

Status: Open.

#120 In swc-17/SparseDrive;

· ioou opened on Nov 17, 2025

How is the closed-loop experiment of SparseDrive implemented

Status: Open.

#119 In swc-17/SparseDrive;

· 13596673883 opened on Sep 25, 2025