FastRCNNConfig¶

class lucid.models.FastRCNNConfig(backbone: lucid.nn.module.Module, feat_channels: int, num_classes: int, pool_size: tuple[int, int] | list[int] = (7, 7), hidden_dim: int = 4096, bbox_reg_means: tuple[float, ...] | list[float] = (0.0, 0.0, 0.0, 0.0), bbox_reg_stds: tuple[float, ...] | list[float] = (0.1, 0.1, 0.2, 0.2), dropout: float = 0.5, proposal_generator: Callable[..., lucid._tensor.tensor.Tensor] | None = None)¶

FastRCNNConfig stores the shared backbone, RoI pooling shape, fully connected head width, bounding-box normalization constants, dropout, and proposal generator used by lucid.models.FastRCNN.

Class Signature¶

@dataclass
class FastRCNNConfig:
    backbone: nn.Module
    feat_channels: int
    num_classes: int
    pool_size: tuple[int, int] | list[int] = (7, 7)
    hidden_dim: int = 4096
    bbox_reg_means: tuple[float, ...] | list[float] = (0.0, 0.0, 0.0, 0.0)
    bbox_reg_stds: tuple[float, ...] | list[float] = (0.1, 0.1, 0.2, 0.2)
    dropout: float = 0.5
    proposal_generator: Callable[..., Tensor] | None = None

Parameters¶

backbone (nn.Module): Shared full-image feature extractor.
feat_channels (int): Channel width of the backbone feature map.
num_classes (int): Number of classification logits produced per RoI.
pool_size: Output size used by RoIAlign.
hidden_dim (int): Hidden width of the two-layer detection head.
bbox_reg_means, bbox_reg_stds: Bounding-box regression normalization constants.
dropout (float): Dropout probability used in the detection head.
proposal_generator: Optional callable that returns pixel-space region proposals.

Validation¶

backbone must be an nn.Module.
feat_channels, num_classes, and hidden_dim must be greater than 0.
pool_size must contain exactly two positive integers.
bbox_reg_means must contain exactly four values.
bbox_reg_stds must contain exactly four positive values.
dropout must be in [0, 1).
proposal_generator must be callable or None.

Usage¶

import lucid.models as models
import lucid.nn as nn

backbone = nn.Sequential(
    nn.Conv2d(3, 64, kernel_size=3, padding=1),
    nn.ReLU(),
)
config = models.FastRCNNConfig(
    backbone=backbone,
    feat_channels=64,
    num_classes=4,
)
model = models.FastRCNN(config)