1 paper · 1 filter
Hao Qu, Lilian Zhang, Jun Mao +5
The performance of visual SLAM in complex, real-world scenarios is often compromised by unreliable feature extraction and matching when using handcrafted features. Although deep le…