The DOI matches its Crossref record; 56 of the 56 references it cites were found in a registry or an index.
FiLM: Visual Reasoning with a General Conditioning Layer
Open accessAbstract
We introduce a general-purpose conditioning method for neural networks called FiLM: Feature-wise Linear Modulation. FiLM layers influence neural network computation via a simple, feature-wise affine transformation based on conditioning information. We show that FiLM layers are highly effective for visual reasoning - answering image-related questions which require a multi-step, high-level process - a task which has proven difficult for standard deep learning methods that do not explicitly model reasoning. Specifically, we show on visual reasoning tasks that FiLM layers 1) halve state-of-the-art error for the CLEVR benchmark, 2) modulate features in a coherent manner, 3) are robust to ablations and architectural modifications, and 4) generalize well to challenging, new data from few examples or even zero-shot.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“While MI-VECTOR (i.e., diagonal parameters in a MI layer which results in a vector output) corresponds to the most basic implementation of FILM, the original FILM layer uses multiple non-linear layers instead of a single linear transformation in MI-VECTOR which has been shown to improve performance [120].”
From MultiBench: Multiscale Benchmarks for Multimodal Representation Learning (2021) · cites this paper as [120]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“By appropriately constraining the rank and structure of these parameters, MI recovers HyperNetworks [61] (unconstrained parameters resulting in a matrix output), Feature-wise linear modulation (FiLM) [120, 188] (diagonal parameters resulting in vector output), and Sigmoid units [37] (scalar parameters resulting in scalar output).”
From MultiBench: Multiscale Benchmarks for Multimodal Representation Learning (2021) · cites this paper as [120]
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
Adam: A Method for Stochastic Optimization
Kingma, D. P., & Ba, J. (2014) · arXiv · doi:10.48550/arxiv.1412.6980 · Cited by 98,946
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380
Squeeze-and-Excitation NetworksPreprint
Hu, J., Shen, L., & Sun, G. (2017) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.48550/arxiv.1709.01507 · Cited by 37,845
Squeeze-and-Excitation Networks
Hu, J., Shen, L., & Sun, G. (2018) · 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2018.00745 · Cited by 35,667
(2008) · Journal of Machine Learning Research · Cited by 33,051
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Ioffe, S., & Szegedy, C. (2015) · arXiv · doi:10.48550/arxiv.1502.03167 · Cited by 27,818
Distributed Representations of Words and Phrases and their Compositionality
Mikolov, T., Sutskever, I., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1310.4546 · Cited by 24,533
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate ShiftPreprint
Ioffe, S. (2024) · arXiv (Cornell University) · doi:10.57702/o9raffed · Cited by 15,273
Research works printed but not linked: 3
- [1][1] A. Krizhevsky, I. Sutskever, and G. E. Hinton, 'Imagenet classification with deep convolutional neural networks,' in Proc. of NIPS , 2012. -
- [4][4] I. Sutskever, O. Vinyals, and Q. V. Le, 'Sequence to sequence learning with neural networks,' in Proc. of NIPS , 2014. -
- [33][33] S. Reed, A. van den Oord, N. Kalchbrenner, V. Bapst, M. Botvinick, and N. de Freitas, 'Generating interpretable images with controllable structure,' in Proc. of ICLR , 2017.
Cited by
Sorted by citations
Items 1,001–1,007 of 1,007
Neural Textual Features Composition for CBIR
Aboali, M. A., El-Maddah, I. A. M., & Munim, H. E. A. E. (2023) · IEEE Access · doi:10.1109/access.2023.3259737 · Cited by 10
Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives
Grauman, K., Westbury, A., Torresani, L., Kitani, K., Malik, J., Afouras, T., Ashutosh, K., Baiyya, V., Bansal, S., Boote, B., et al. (2025) · International Journal of Computer Vision · doi:10.1007/s11263-025-02557-6 · Cited by 3
Controllable Single-Shot Animation Blending with Temporal Conditioning
Tselepi, E., Thermos, S., & Potamianos, G. (2025) · 2025 IEEE/CVF International Conference on Computer Vision Workshops (ICCVW) · doi:10.1109/iccvw69036.2025.00050 · Cited by 1
Tiwari, K., Mahajan, G., Bellad, R., Dhruw, N. K., & Jha, A. (2026) · doi:10.64898/2026.10.16.759604
Cross-robot behavior adaptation through intention alignment
Chen, X., Gao, Y., Liu, H., Yang, F., Ghadirzadeh, A., Yang, J., Liang, B., Zhang, C., Lam, T. L., & Zhu, S.-C. (2026) · Science Robotics · doi:10.1126/scirobotics.adv2250
Language as an Abstraction for Hierarchical Deep Reinforcement Learning
Jiang, Y., Gu, S., Murphy, K., & Finn, C. (2019) · arXiv · doi:10.48550/arxiv.1906.07343 · Cited by 109
PHYRE: A New Benchmark for Physical Reasoning
Bakhtin, A., van der Maaten, L., Johnson, J., Gustafson, L., & Girshick, R. (2019) · arXiv · doi:10.48550/arxiv.1908.05656 · Cited by 37
Metadata
- Authors
- Perez, E., Strub, F., De Vries, H., Dumoulin, V., & Courville, A.
- Affiliations
- Université de Montréal (CA)
- Centre National de la Recherche Scientifique (FR)
- Mila - Quebec Artificial Intelligence Institute (CA)
- Venue
- Proceedings of the AAAI Conference on Artificial Intelligence
- Published
- 2018-04-29
- Volume
- 32
- Issue
- 1
- ISSN
- 2159-5399 (ISSN-L)
- 2374-3468
- OpenAlex
- W2760103357
- Cited by
- 2,490
- References
- 55
- Citations by year
- 2026: 823 · 2025: 412 · 2024: 280 · 2023: 221 · 2022: 158 · 2021: 244 · 2020: 179 · 2019: 120 · 2018: 52
- FWCI
- 44.92
- Type
- conference paper
- Language
- en
- Open access
- diamond
- doi.org
- Locations
- Proceedings of the AAAI Conference on Artificial Intelligence (publishedVersion)
- arXiv (Cornell University) (submittedVersion)
- AAAI Publications (The Association for the Advancement of Artificial Intelligence (AAAI)) (submittedVersion)
- Funders
- Nvidia
- Fonds de recherche du Québec – Nature et technologies
- CHIST-ERA
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Multimodal Machine Learning Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- cinema, feature-wise linear modulation, visual reasoning, visual question answering, feature modulation, zero-shot generalization, few-shot learning
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
