![[2602.13298] Effect of Convolutional Depth on Image Recognition Performance: VGG vs. ResNet vs. GoogLeNet](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.13298] Effect of Convolutional Depth on Image Recognition Performance: VGG vs. ResNet vs. GoogLeNet
Summary
This paper examines how convolutional depth affects image recognition performance across three architectures: VGG, ResNet, and GoogLeNet, revealing that effective depth is more crucial than nominal depth for achieving better accuracy.
Why It Matters
Understanding the relationship between convolutional depth and image recognition performance is vital for optimizing neural network architectures. This study provides insights that can guide researchers and practitioners in selecting and designing more efficient models, ultimately enhancing the effectiveness of AI applications in computer vision.
Key Takeaways
- Effective depth, rather than nominal depth, is key to improving accuracy in CNNs.
- Residual and inception-based architectures outperform plain deep networks in terms of optimization stability.
- Increased depth does not guarantee better performance; architectural mechanisms play a critical role.
- Standardized training protocols can reveal the true impact of depth on performance.
- The study highlights the importance of understanding architectural constraints in deep learning.
Computer Science > Computer Vision and Pattern Recognition arXiv:2602.13298 (cs) [Submitted on 9 Feb 2026] Title:Effect of Convolutional Depth on Image Recognition Performance: VGG vs. ResNet vs. GoogLeNet Authors:Manfred M. Fischer, Joshua Pitts View a PDF of the paper titled Effect of Convolutional Depth on Image Recognition Performance: VGG vs. ResNet vs. GoogLeNet, by Manfred M. Fischer and Joshua Pitts View PDF HTML (experimental) Abstract:Increasing convolutional depth has been central to advances in image recognition, yet deeper networks do not uniformly yield higher accuracy, stable optimization, or efficient computation. We present a controlled comparative study of three canonical convolutional neural network architectures - VGG, ResNet, and GoogLeNet - to isolate how depth influences classification performance, convergence behavior, and computational efficiency. By standardizing training protocols and explicitly distinguishing between nominal and effective depth, we show that the benefits of depth depend critically on architectural mechanisms that constrain its effective manifestation during training rather than on nominal depth alone. Although plain deep networks exhibit early accuracy saturation and optimization instability, residual and inception-based architectures consistently translate additional depth into improved accuracy at lower effective depth and favorable accuracy-compute trade-offs. These findings demonstrate that effective depth, not nominal depth...
