各CF-based tracker中output_sigma_factor取值
现有的各CF-Based tracker中理想高斯响应中output_sigma_factor的取值情况
默认output_sigma = target_sz*output_sigma_factor;
【CREST】 0.1
output_sigma_factor=0.1;
output_sigma = target_sz*output_sigma_factor;%output_sigma 是个向量
labels = exp(-alpha*(rs.^2/sigma(1)^2 + cs.^2/sigma(2)^2));%这里alpha根据不同的情况取0.2或者0.3,具体参见源代码
【KCF】0.1
output_sigma_factor=0.1;
output_sigma = sqrt(prod(target_sz)) * output_sigma_factor / cell_size; %output_sigma 是个标量
labels = exp(-0.5 / sigma^2 * (rs.^2 + cs.^2));
Reference:
【CREST】:Y Song, C Ma, L Gong, et al. CREST Convolutional Residual Learning for Visual Tracking[C]. //ICCV2017
【KCF】: JF Henriques, R Caseiro, P Martins, et al. High-Speed Tracking with Kernelized Correlation Filters[J]. TPAMI2015.