目录

目录

Differentiable Volumetric Rendering: Learning Implicit 3D Representations without 3D Supervision


<DVR> Differentiable volumetric rendering: Learning implicit 3d representations without 3d supervision
目录

review

  • 思路:首先手动推导出每个camera ray和隐表面交点的点坐标对网络参数的梯度,在实际计算时,就可以先在camera ray上采样得出交点坐标(类似二分法),然后代入所手动推导出的式子构成完整的反向传播链路