Good View Hunting: Learning Photo Composition from Dense View Pairs

Listwise View Ranking for Image Cropping

“…All weights of the three FC layers are initialized with normal distribution (zero mean and 0.01 standard deviation), bias are set to zero and the channels are set to 1024, 512 and 1, respectively. The proposed model is trained on CPC dataset [Wei et al, 2018] including 10,797 images, each with 24 candidate views. We directly rank 24 views and assign the order of views as ground-truth rank score.…”

“…IoU Avg. Disp AesRankNet [Kong et al, 2016] 0.4843 0.1400 RankSVM [Chen et al, 2017a] 0.6020 0.1060 VFN+SW [Chen et al, 2017b] 0.6328 0.0982 A2-RL [Li et al, 2018] 0.6633 0.0892 VPN [Wei et al, 2018] 0.6641 0.0858 LVRN (Ours) 0.7100 0.0735 choose the views with best rank score f v j i . Here we set the size of search windows among [0.6, 0.65, 0.7, ..., 0.9] and the aspect ratio among [1:1, 3:4, 4:3, 9:16, 16:9].…”

“…FCDB contains 348 test images and each image is labelled by a photography hobbyist. To evaluate the generalization ability of our model, we adopt the same metrics as previous works [Chen et al, 2017a;Wei et al, 2018], including intersectionover-union (IoU) and boundary displacement (Disp). The IoU can be computed as…”

“…With simple pair selection, we set a threshold 0.5 to drop the pairs with a minor gap of rank score, and generate 1.3 million pairs. With careful pair selection, we train the model following [Wei et al, 2018]. The markedly improvement in Table 5 shows that listwise training overcomes the problem of the pairwise training.…”

“…View finding network (VFN) [Chen et al, 2017b] based on a pairwise ranking layer is proposed to model the photo composition and crop image by sliding window. Wei et al trained a view evaluation network (VEN) [Wei et al, 2018] with the pairwise siamese architecture. Inspired by knowledge distillation [Hinton et al, 2015] and anchor boxes [Liu et al, 2016], view proposal network (VPN) is proposed to transfer knowledge from VEN.…”