Open Problems

Multimodal Fusion & Multi-Task Learning

Empirical Robustness and Scalability Benchmarking for Multi-View Multimodal Fusion

Scope to testOpen
Strong candidate · 5/5 runs8 papers report this62% from 2025+

Generated automatically from the limitations stated in 8 papers (CVPR, NeurIPS, ICLR), listed under Evidence. It is not a paper, and it does not come from papers submitted to CSPaper.

The problem

Current multi-view and multimodal fusion methods are evaluated almost exclusively on small benchmarks (typically under 10,000 samples) with very few modalities or views (2 to 6 views). Because scaling behavior across large sample counts ($N \gg 10^4$) and high view counts ($V > 6$) remains untested, it is unknown whether reported algorithmic advantages persist, collapse, or become computationally intractable in larger settings. Practitioners cannot reliably select fusion algorithms for complex multi-sensor or high-view applications because no empirical baseline exists across these broader operating conditions.

Why it matters

Provides the first empirical ground truth on how multi-view fusion algorithms perform and scale when pushed beyond small benchmark limits. Enables practitioners to select robust fusion architectures for large-scale, high-view deployments based on demonstrated empirical behavior.

Ways to approach it

Prior-work checks are free with an account. Results someone already ran are shown to everyone.

  1. 1

    Unified Scaling Testbed: Construct a standardized evaluation pipeline containing top-performing multi-view fusion paradigms (e.g., cross-view alignment, tensor decomposition, and deep shared-space encoders) and evaluate them on datasets scaled from $V=2$ to $V=30$ views and $N=10^3$ to $N=500,000$ samples. Measure task accuracy (classification/clustering), memory footprints, and wall-clock training time against theoretical complexity bounds.

  2. 2

    View-Robustness and Missing-View Stress Testing: Systematically vary the number of available views and view-to-sample ratios on high-dimensional benchmarks to evaluate model stability. Measure the performance delta and ranking stability of methods as view count scales upward.

Have a different approach?

Describe how you would tackle this problem and we'll look for papers that already do it. Free; your text stays private.

Free · 3 checks per day

Why it might fail

If standard baseline implementations encounter out-of-memory errors that prevent running full multi-view baselines without bespoke algorithmic re-engineering, completing the benchmark matrix may prove intractable. Additionally, if the scaling behavior across all methods simply mirrors trivial theoretical parameter complexity without unexpected performance shifts, the empirical contribution would be marginal.

Evidence

Each paper's own statement of the limitation, verbatim.

Nearest existing work

Related open problems

Multimodal Fusion & Multi-Task Learning

Barrier to removePartly addressed

Systematic Robustness Benchmarking of Incomplete Multimodal Learning Under Non-Random and Structured Missingness

Existing multimodal learning methods designed for missing data are predominantly evaluated under Missing at Random (MAR) or uniform independent drop assumptions, often requiring fully complete training data. In practical applications like clinical diagnostics and robotics, missingness is frequently structured, correlated across sensors, or Missing Not at Random (MNAR) due to cost or clinical decisions. Because existing literature only tests idealized missingness or uses simple mean imputation fallbacks, practitioners cannot determine whether current alignment, distillation, or generative imputation methods generalize to realistic incomplete data regimes.

Possible candidate · 2/5 runs9 papers report this67% from 2025+
Generated automatically, not curated by hand. Automated prior-work checks catch about a third of existing work, so treat this problem as a lead to investigate.