ImageStar & VolumeStar¶
ImageStar¶
The ImageStar extends the Star set concept to multi-channel 2D images, preserving spatial structure through convolutional layers.
Definition¶
where:
\(c \in \mathbb{R}^{H \times W \times C}\) is the center image
\(v_i \in \mathbb{R}^{H \times W \times C}\) are generator images
\(\alpha \in \mathbb{R}^m\) are predicate variables
\(C\alpha \leq d\) are linear constraints
Key Insight¶
The ImageStar preserves the spatial structure of images, which is critical for efficient convolution. Rather than flattening to a 1D Star (which would make convolution a dense matrix multiply), the ImageStar applies convolution directly to the center and each generator image:
This means the convolution of an ImageStar is computed as:
Convolve the center image
Convolve each generator image independently
The constraints remain unchanged
This is vastly more efficient than converting to a dense Star representation.
Operations¶
Operation |
ImageStar Behavior |
|---|---|
Convolution (Conv2D) |
Apply conv to center and each generator |
Batch normalization |
Scale and shift center and generators |
Average pooling |
Apply pooling to center and generators |
Max pooling |
Requires LP-based splitting or approximation |
ReLU |
Per-pixel exact or approximate analysis |
Flatten |
Convert to 1D Star set |
VolumeStar¶
The VolumeStar further extends ImageStar to 4-dimensional data: 3D spatial volumes or video frames.
Definition¶
where:
\(c \in \mathbb{R}^{H \times W \times C \times F}\) is the center volume
\(v_i \in \mathbb{R}^{H \times W \times C \times F}\) are generator volumes
The 4th dimension \(F\) can represent depth (for 3D medical images) or frames (for video)
Supported Layers¶
The VolumeStar propagates through:
Conv3D: 3D convolution applied to center and generators
AveragePooling3D: 3D spatial pooling
ReLU / activation layers: Per-voxel analysis
Flatten: Convert to 1D Star
Applications¶
Video classification: Verify C3D/I3D networks under temporal and spatial perturbations
3D medical imaging: Verify classifiers on MRI/CT volumes under noise and artifact perturbations
Spatio-temporal robustness: Perturbations that affect both space and time dimensions
Hierarchy¶
The set representations form a natural hierarchy:
Star (1D vectors)
└── ImageStar (2D: H × W × C)
└── VolumeStar (3D: H × W × C × F)
Zono (1D zonotopes)
└── ImageZono (2D zonotopes)
GraphStar (graph-structured: N × F)
Each level specializes the parent representation for higher-dimensional data while maintaining the same constraint structure (\(C\alpha \leq d\)).
References¶
H.-D. Tran, S. Bak, W. Xiang, T.T. Johnson, “Towards Verification of Large Convolutional Neural Networks Using ImageStars,” CAV 2020
S. Sasaki, D. Manzanas Lopez, P.K. Robinette, T.T. Johnson, “Robustness Verification of Video Classification Neural Networks,” FormaliSE 2025