diff --git a/luma/neural/block.py b/luma/neural/block.py index b22b499..9c54cc6 100644 --- a/luma/neural/block.py +++ b/luma/neural/block.py @@ -560,6 +560,25 @@ class IncepBlockArgs: @ClassType.non_instantiable() class IncepBlock: + """ + Container class for various Inception blocks. + + References + ---------- + Inception V1 & V2 : + [1] Szegedy, Christian, et al. “Going Deeper with Convolutions.” + Proceedings of the IEEE Conference on Computer Vision and + Pattern Recognition (CVPR), 2015, pp. 1-9, + arxiv.org/abs/1409.4842. + + Inception V4, Inception-ResNet V1 & V2: + [2] Szegedy, Christian, et al. “Inception-v4, Inception-ResNet + and the Impact of Residual Connections on Learning.” + Proceedings of the Thirty-First AAAI Conference on + Artificial Intelligence (AAAI), 2017, pp. 4278-4284, + arxiv.org/abs/1602.07261. + + """ class V1(_specials.incep_v1._Incep_V1_Default): """ @@ -567,11 +586,7 @@ class V1(_specials.incep_v1._Incep_V1_Default): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Going Deeper with Convolutions.” - Proceedings of the IEEE Conference on Computer Vision and - Pattern Recognition (CVPR), 2015, pp. 1-9, arxiv.org/abs/1409.4842. + Refer to the figures shown in the original paper[1]. """ class V2_TypeA(_specials.incep_v2._Incep_V2_TypeA): @@ -580,12 +595,7 @@ class V2_TypeA(_specials.incep_v2._Incep_V2_TypeA): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Going Deeper with Convolutions.” - Proceedings of the IEEE Conference on Computer Vision and - Pattern Recognition (CVPR), 2015, pp. 1-9, - arxiv.org/abs/1409.4842. + Refer to the figures shown in the original paper[1]. """ class V2_TypeB(_specials.incep_v2._Incep_V2_TypeB): @@ -594,12 +604,7 @@ class V2_TypeB(_specials.incep_v2._Incep_V2_TypeB): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Going Deeper with Convolutions.” - Proceedings of the IEEE Conference on Computer Vision and - Pattern Recognition (CVPR), 2015, pp. 1-9, - arxiv.org/abs/1409.4842. + Refer to the figures shown in the original paper[1]. """ class V2_TypeC(_specials.incep_v2._Incep_V2_TypeC): @@ -608,12 +613,8 @@ class V2_TypeC(_specials.incep_v2._Incep_V2_TypeC): Parameters ---------- - Refer to the figures shown in the original paper: + Refer to the figures shown in the original paper[1]. - Szegedy, Christian, et al. “Going Deeper with Convolutions.” - Proceedings of the IEEE Conference on Computer Vision and - Pattern Recognition (CVPR), 2015, pp. 1-9, - arxiv.org/abs/1409.4842. """ class V2_Redux(_specials.incep_v2._Incep_V2_Redux): @@ -622,12 +623,8 @@ class V2_Redux(_specials.incep_v2._Incep_V2_Redux): Parameters ---------- - Refer to the figures shown in the original paper: + Refer to the figures shown in the original paper[1]. - Szegedy, Christian, et al. “Going Deeper with Convolutions.” - Proceedings of the IEEE Conference on Computer Vision and - Pattern Recognition (CVPR), 2015, pp. 1-9, - arxiv.org/abs/1409.4842. """ class V4_Stem(_specials.incep_v4._Incep_V4_Stem): @@ -636,13 +633,7 @@ class V4_Stem(_specials.incep_v4._Incep_V4_Stem): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Inception-v4, Inception-ResNet - and the Impact of Residual Connections on Learning.” - Proceedings of the Thirty-First AAAI Conference on - Artificial Intelligence (AAAI), 2017, pp. 4278-4284, - arxiv.org/abs/1602.07261. + Refer to the figures shown in the original paper[2]. Notes ----- @@ -660,13 +651,7 @@ class V4_TypeA(_specials.incep_v4._Incep_V4_TypeA): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Inception-v4, Inception-ResNet - and the Impact of Residual Connections on Learning.” - Proceedings of the Thirty-First AAAI Conference on - Artificial Intelligence (AAAI), 2017, pp. 4278-4284, - arxiv.org/abs/1602.07261. + Refer to the figures shown in the original paper[2]. Notes ----- @@ -684,13 +669,7 @@ class V4_TypeB(_specials.incep_v4._Incep_V4_TypeB): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Inception-v4, Inception-ResNet - and the Impact of Residual Connections on Learning.” - Proceedings of the Thirty-First AAAI Conference on - Artificial Intelligence (AAAI), 2017, pp. 4278-4284, - arxiv.org/abs/1602.07261. + Refer to the figures shown in the original paper[2]. Notes ----- @@ -708,13 +687,7 @@ class V4_TypeC(_specials.incep_v4._Incep_V4_TypeC): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Inception-v4, Inception-ResNet - and the Impact of Residual Connections on Learning.” - Proceedings of the Thirty-First AAAI Conference on - Artificial Intelligence (AAAI), 2017, pp. 4278-4284, - arxiv.org/abs/1602.07261. + Refer to the figures shown in the original paper[2]. Notes ----- @@ -733,14 +706,8 @@ class V4_ReduxA(_specials.incep_v4._Incep_V4_ReduxA): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Inception-v4, Inception-ResNet - and the Impact of Residual Connections on Learning.” - Proceedings of the Thirty-First AAAI Conference on - Artificial Intelligence (AAAI), 2017, pp. 4278-4284, - arxiv.org/abs/1602.07261. - + Refer to the figures shown in the original paper[2]. + Notes ----- - This block has fixed shape of input and ouput tensors. @@ -758,13 +725,7 @@ class V4_ReduxB(_specials.incep_v4._Incep_V4_ReduxB): Parameters ---------- - Refer to the figures shown in the original paper: - - Szegedy, Christian, et al. “Inception-v4, Inception-ResNet - and the Impact of Residual Connections on Learning.” - Proceedings of the Thirty-First AAAI Conference on - Artificial Intelligence (AAAI), 2017, pp. 4278-4284, - arxiv.org/abs/1602.07261. + Refer to the figures shown in the original paper[2]. Notes -----