High-quality Strong Lens Candidates in the Final Kilo-Degree Survey Footprint

We present 97 new high-quality strong lensing candidates found in the final $\sim 350\,\rm deg^2$, that completed full 1350\,\rm deg^2$ area of Kilo-Degree Survey (KiDS). Together with our previous findings, list from KiDS sums up to 268 systems. The sample is assembled using a Convolutional Neural Network (CNN) classifier applied $r$-band (best seeing) and $g,~r,~i$ color-composited images separately. This optimizes complementarity morphology color information on identification candidates. apply classifiers luminous red galaxies (LRGs) bright (BGs) select received high probability be lens CNN ($P_{\rm CNN}$). In particular, setting $P_{\rm CNN}>0.8$ for LRGs, $1$-band predicts 1213 candidates, while $3$-band yields 1299 only $\sim$30\% overlap. For BGs, order minimize false positives, we adopt more conservative threshold, CNN} >0.9$, both classifiers. results 3740 newly selected objects. two samples are visually inspected by 7 co-authors finally "high-quality" which mean scores larger than 6 (on scale 0 10). discuss effect seeing accuracy classification possible avenues increase efficiency multi-band classifiers, preparation next-generation surveys ground space.