Moviescope: Large-scale Analysis of Movies using Multiple Modalities

Kalpathy Sitaraman; Mengjia Luo; Paola Cascante-Bonilla; Vicente Ordonez

REVIEW 4 major objections 6 minor 39 references

Moviescope: Large-scale Analysis of Movies using Multiple Modalities

T0 review · 4 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that simple average-pooled pretrained frame features outperform LSTM, C3D, and I3D models for predicting movie genres and budgets.

desk verdict Moviescope is a genuinely useful multimodal movie benchmark with a plausible finding that simple pooling beats recurrent models, but the data alignment is under-verified and the dataset is only promised for release, so the paper needs revision before the claims are fully acceptable. read the letter →

arxiv 1908.03180 v1 pith:QE4AX6PC submitted 2019-08-08 cs.CV

Paola Cascante-Bonilla , Kalpathy Sitaraman , Mengjia Luo , Vicente Ordonez This is my paper · ORCID

classification cs.CV

keywords movieunderstandingmultimodallearningvideorepresentationaveragepoolinggenreclassificationbudgetestimationtraileranalysisdataset

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper sets out to show that for understanding movies at the level of genre and budget, simple content-based representations can beat more complex temporal models, and that different modalities carry complementary signals. It introduces Moviescope, a dataset of more than 5,000 movies with trailers, plots, posters, and metadata, and benchmarks several encodings. The central finding is that fastVideo and fastText, which average frame and word features, outperform LSTM, C3D, and I3D baselines on these holistic movie tasks. If the finding holds, average pooling with pretrained features is a strong, cheap baseline for other long-form video classification tasks.

What carries the argument

The central objects are fastText and its video analogue fastVideo: representations built by feeding pretrained per-word or per-frame features through an affine layer after a global average pooling operation over the sequence. For video, frames are subsampled to 200 per trailer and encoded with the penultimate layer of VGG-16; for text, plots up to 3,000 words are encoded with GloVe embeddings. This pooling operation replaces explicit temporal modeling, and it is what allows the paper to compare cheap holistic encodings against LSTM, C3D, and I3D on holistic movie-level tasks.

What would settle it

Take a random sample of about 200 Moviescope entries, manually check whether the downloaded trailer and parsed plot actually correspond to the IMDb movie, and record the mismatch rate; then rerun the fastVideo versus LSTM genre comparison on only verified pairs and see whether the reported advantage disappears.

Watch

Extended reading notes

Core claim

Moviescope is a new multimodal dataset of 5,027 movies, with YouTube trailers (video plus audio), Wikipedia plot summaries, posters, and IMDb-derived metadata. On this data, for multilabel genre prediction over 13 non-exclusive genres and for 5-tier budget estimation, the paper's fastVideo encoding—average pooling over VGG-16 frame activations—achieves higher mean average precision than LSTM, bidirectional LSTM, C3D, and Two-Stream I3D baselines. fastText, the analogous average pooling over GloVe word embeddings, is the strongest single modality for genre, and audio and text outperform video for budget. Fusing all modalities with learned per-genre attention weights yields the best overall scores.

Load-bearing premise

The automatically downloaded first YouTube result for each movie-title-plus-trailer query and the regex-parsed Wikipedia plot are correctly matched to the corresponding IMDb movie record, with no manual verification of the alignment for the full dataset.

Editorial extensions

If this is right

Simple pooling with pretrained features should be the default baseline for movie-level semantic tasks, since it beats recurrent and 3D convolutional models at lower computational cost.
Trailers carry enough information about a full movie to predict genre, so trailer-level content can support recommendation and retrieval without full-length video analysis.
Audio is a stronger budget signal than video, suggesting production scale is audible in sound design and mixing.
Multimodal fusion with per-genre attention improves over any single modality, and the learned attention weights reveal which signal matters most for each genre.
The dataset provides a new benchmark for joint video, text, audio, and image understanding, with aligned modalities for the same movies.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Editorial inference: the same average-pooling recipe may transfer to other long-form media tasks such as TV episode classification, because the paper's evidence suggests temporal order is not crucial for holistic labels.
The dataset's alignment quality is the key risk; a small manual audit of trailer–plot–movie matches would quantify the mismatch rate and could be reported alongside the benchmark.
The modality attention weights suggest a content-based recommender could emphasize trailers for animation, plots for drama, and audio for high-budget films—an application the paper does not build out.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Here's my take: the paper's real contribution is the dataset concept—5k movies with trailers, plots, posters, and metadata in a 13-way multi-label genre scheme—and the empirical claim that average-pooled pretrained features (fastText/fastVideo) beat LSTM and 3D models on movie-level tasks. That claim is plausible and the paper gives it a fair initial test, including held-out evaluation and a human study. It deserves a serious referee, but not acceptance in this form.

What it does well: the experimental structure is clear. They compare five modalities, build a fusion with attention weights, and check against human workers. The finding that text and audio are better for budget than video is non-obvious. The budget tiers are sensible. The UCF101 table is labeled honestly: fastVideo is not SOTA, and they say so.

The soft spots are real. The data collection in Section 3 is under-specified: 'first YouTube result' plus regex-parsed Wikipedia plots, with no manual verification or quality metric. The intro and related work claim manual curation, but the collection section never describes that manual step. That discrepancy matters for a dataset paper. If a nontrivial share of trailers or plots are misaligned, the benchmark numbers are tainted. The effect may not be random—older and obscure films are more likely to fetch wrong videos. The authors should add a verification step or soften the guarantee.

The other issues are more moderate. There are no error bars or significance tests, so the margin of fastVideo over C3D (56.5 vs 53.4 mAP) may be noise. The LSTM clip lengths (16/49 frames) were chosen on validation, which is fine, but they don't report variance across splits. The UCF101 comparison uses published numbers without a controlled rerun, so it's context, not evidence. And the dataset is only promised, not released—the abstract and intro say 'plan to release,' so the reproducibility claims are forward-looking.

Bottom line: the core idea is sound and the paper is a useful benchmark contribution for anyone working on movie understanding or multimodal analysis. But the data integrity concern is the linchpin, and the lack of release and error bars make the result conditional. I'd send it to review, asking for verification of the alignments, error bars, and a clear release commitment.

Referee Report

4 major / 6 minor

Summary. The paper introduces Moviescope, a multimodal dataset of about 5,000 movies with YouTube trailers (video and audio), Wikipedia-derived plots, posters, and IMDb-based metadata, and uses it to study multi-label genre classification and budget estimation. It proposes fastVideo, which average-pools VGG-16 frame features over a subsampled sequence, as a video analog of fastText for text. Experiments compare fastVideo against LSTM, bidirectional LSTM, C3D, and I3D for video; fastText and LSTM for text; a CRNN for audio; VGG-16 for posters; and random forest/XGBoost for metadata, as well as score fusion. The central empirical claim is that simple pooling of pretrained features outperforms recurrent and 3D-convolutional baselines for both tasks, and that text and audio are strong predictors. The paper also reports a human study for genre prediction and comparisons on UCF101.

Significance. If the empirical comparisons hold, the paper makes a useful contribution: it provides a new multimodal benchmark, demonstrates that average-pooled pretrained features are a strong and cheap baseline for movie-level semantics, and identifies several non-obvious findings (e.g., audio predicts budget better than video). The planned release of the dataset, pretrained embeddings, and code would increase reproducibility. However, the value of the benchmark and the validity of the central comparison depend critically on the integrity of the crawled trailer-plot-movie alignments and on fair, controlled baselines, both of which are currently insufficiently documented.

major comments (4)

[Section 3 and Introduction] The paper claims that the pipeline 'guarantee[s] downloading the best possible video trailer' and that the release includes 'manually curated mappings between these two modalities,' but Section 3 describes only fetching the first YouTube search result for each query and regex-parsing Wikipedia HTML, with no manual verification, agreement metric, or quality-control threshold. The 4,996 automatically downloaded trailers and the parsed plots are not checked against the IMDb records (e.g., for title/year match). This is load-bearing because every content-based result in Tables 3 and 5 (including the fastVideo vs. LSTM/C3D/I3D comparison) is computed on these pairings; if even a modest fraction are wrong, the rankings can change. The authors should report a validation of the alignments, for instance a manual audit on a random sample with an estimated mismatch rate.
[Section 6, Table 3 and LSTM baseline paragraph] The paper states that 'choosing random clips of 16 or 49 continuous frames worked better than using 200 frames at once' and that at test time the LSTM sees 12 clips of 16 frames and 4 clips of 49 frames, which are then averaged. This protocol prevents the LSTM from ever seeing the full temporal extent of the trailer, while fastVideo averages over 200 frames drawn from the whole trailer. The comparison therefore conflates the aggregation method with the amount of temporal context available. The authors should either provide LSTM results on the full 200-frame sequence (with comparable capacity) or justify why the clip-based LSTM is the appropriate strong baseline; otherwise the central claim that fastVideo 'provides higher levels of accuracy' is not fully supported.
[Section 6, Table 4 (UCF101)] The table is titled 'Mean Average Precision Scores on UCF101,' but the baseline numbers cited from [16,34,37,28,5] are top-1 classification accuracies reported in the original papers (UCF101 is a single-label dataset). Comparing these published accuracy numbers to the mAP numbers reported for fastVideo is an uncontrolled, apples-to-oranges comparison. The authors should reimplement the baselines under the same evaluation protocol (same metric, same input frames, same downstream classifier) or at minimum disclose the metric mismatch and limit conclusions accordingly. As written, the statement that fastVideo 'does outperform similar competing approaches such as LSTM encodings or C3D' on UCF101 is not justified.
[Tables 3 and 5, Section 6] All results are reported as single mAP values from one random 70/10/20 split, with no error bars, confidence intervals, or significance tests. Given that several comparisons are close (e.g., fastVideo vs. C3D on several genres in Table 3, and I3D RGB vs. fastVideo on budget tiers in Table 5), the reader cannot determine whether the observed differences are statistically reliable. The authors should report results over multiple seeds and/or multiple splits, and perform a significance test (e.g., paired bootstrap) for the headline fastVideo vs. LSTM/C3D/I3D differences.

minor comments (6)

[Section 3 and Table 1] The number of movies is stated as 5,000 in the abstract, 5,043 in Section 3 ('IMDb5000 dataset consisting of 5,043 movie records'), 4,996 after trailer download, and 4,927 in training after discarding those without trailer/plot; please make the dataset size consistent and clarify the exact splits.
[Figure 1 caption] The caption contains the typo 'previoulsy' and should read 'previously.'
[Section 4.1] The word 'enconding' should be 'encoding.'
[Section 6] The text contains the typos 'Morever' (should be 'Moreover') and 'cleary' (should be 'clearly').
[Section 4.2] The sentence about supplementing frames ('starting at the 200th frame by taking one every 6 frames') is confusing; specify whether the second pass starts after the first 200 sampled frames or at the 200th frame of the trailer.
[Equations (1) and (2)] The summation indices start at j=0 and i=0, but the text refers to word/frame embeddings starting from the first element; use j=1 and i=1 or define the zero-index convention.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: all content-based predictions are evaluated on held-out test splits using features from externally pretrained models, with no target-label leakage into the inputs.

full rationale

The paper's central claims are empirical comparisons on held-out test splits. The fastVideo, fastText, poster, and audio representations are built from VGG-16 (ImageNet-pretrained), GloVe (Wikipedia-pretrained), and spectral features, none of which are trained on the movie genre or budget labels. The task-specific affine layers and fusion weights are fitted on the training split and evaluated on a separate test split, which is standard supervised model fitting rather than circular reasoning. No equation in the paper defines a predicted quantity in terms of the target label, and no fitted parameter is renamed as a prediction. The only load-bearing external references are standard models (C3D, I3D, LSTM, fastText, GloVe, VGG-16), and the paper additionally benchmarks fastVideo on the external UCF101 dataset, providing independent empirical anchoring. The concerns about unverified YouTube 'first search result' trailer alignment and regex-parsed Wikipedia plots are data-quality and correctness risks, not circularity: even if misalignment exists, that would invalidate the data rather than make the derivation equivalent to its inputs. No self-citation chain is used to justify the main result. Therefore the derivation is self-contained with respect to circularity.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central experiments rest on standard supervised model fitting plus several hand-chosen design parameters (frame counts, budget tiers, audio clip sizes, clip lengths for LSTM baselines). The dataset curation depends on the domain assumption that automatically fetched YouTube trailers and Wikipedia plots are correctly aligned with IMDb records. No new physical or conceptual entities are introduced.

free parameters (5)

LSTM clip lengths (16/49 frames) = 16 and 49 frames
Chosen after validation experiments (Section 6) and used for LSTM comparisons, making the LSTM baseline weaker than if full 200-frame context were used.
Budget tier boundaries = 5 tiers: $218-890K, $900K-4.8M, $4.9M-19.4M, $19.5M-71.5M, $72M-300M
Ad hoc discretization in Section 6.1; mAP depends on these boundaries and they are not derived from data.
Temporal sampling rate for fastVideo = 200 frames, one every 10 frames; fallback every 6 frames
Chosen in Section 4.2; no sensitivity analysis is reported.
Audio clip configuration = 4 clips of 30 seconds, 128 mel bins, hop size 256
Chosen in Section 4.3; no sensitivity analysis is reported.
Multimodal fusion weights alpha = learned via softmax over per-modality scores
Learned parameters in Eq. 3-4; they are fitted to the same training set and then interpreted as modal attention in Figure 4.

assumptions (4)

domain assumption Video trailers are representative summaries of their full-length movies
Stated in Section 1 and the Introduction; if false, trailer-based genre and budget predictions are invalid.
domain assumption Pre-trained VGG-16 ImageNet features transfer to movie genre and budget prediction
Used as a fixed feature extractor in Sections 4.2 and 4.4; no fine-tuning on movie data is performed.
domain assumption Pre-trained GloVe embeddings transfer to movie plot text
Used in Section 4.1; no task-specific training of word vectors is performed.
domain assumption IMDb genre labels and budget metadata are reliable ground truth
The dataset is built on IMDb5000 (Section 3); label noise is not analyzed.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Moviescope: Large-scale Analysis of Movies using Multiple Modalities." pith.science (2026). https://pith.science/paper/QE4AX6PC

@misc{pith2026190803180,
  author       = {Pith},
  title        = {Pith review of: Moviescope: Large-scale Analysis of Movies using Multiple Modalities},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QE4AX6PC}},
  note         = {Machine review of arXiv:1908.03180}
}

read the original abstract

Film media is a rich form of artistic expression. Unlike photography, and short videos, movies contain a storyline that is deliberately complex and intricate in order to engage its audience. In this paper we present a large scale study comparing the effectiveness of visual, audio, text, and metadata-based features for predicting high-level information about movies such as their genre or estimated budget. We demonstrate the usefulness of content-based methods in this domain in contrast to human-based and metadata-based predictions in the era of deep learning. Additionally, we provide a comprehensive study of temporal feature aggregation methods for representing video and text and find that simple pooling operations are effective in this domain. We also show to what extent different modalities are complementary to each other. To this end, we also introduce Moviescope, a new large-scale dataset of 5,000 movies with corresponding movie trailers (video + audio), movie posters (images), movie plots (text), and metadata.

Figures

Figures reproduced from arXiv: 1908.03180 by the authors.

**Figure 2.** Label statistics in the Moviescope dataset, showing co [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Overview of our feature representations for multimodal [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Attention on different modalities for a select subset of genres. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: Example predictions of the best performing models for each modality. We show some sample frames for the video trailers, and a [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 26 canonical work pages

[1]

Abu-El-Haija, N

S. Abu-El-Haija, N. Kothari, J. Lee, P. Natsev, G. Toderici, B. Varadarajan, and S. Vijayanarasimhan. Youtube-8m: A large-scale video classiﬁcation benchmark. CoRR, abs/1609.08675, 2016

arXiv 2016
[2]

Bamman, B

D. Bamman, B. O’Connor, and N. A. Smith. Learning la- tent personas of ﬁlm characters. In Proceedings of the An- nual Meeting of the Association for Computational Linguis- tics (ACL), page 352, 2014

work page 2014
[3]

B ¨auml, M

M. B ¨auml, M. Tapaswi, and R. Stiefelhagen. Semi- supervised learning with constraints for person identiﬁcation in multimedia data. In Computer Vision and Pattern Recog- nition (CVPR), 2013 IEEE Conference on, pages 3602–3609. IEEE, 2013

work page 2013
[4]

Bojanowski, F

P. Bojanowski, F. Bach, I. Laptev, J. Ponce, C. Schmid, and J. Sivic. Finding actors and actions in movies. In Com- puter Vision (ICCV), 2013 IEEE International Conference on, pages 2280–2287. IEEE, 2013

work page 2013
[5]

Carreira and A

J. Carreira and A. Zisserman. Quo vadis, action recog- nition? A new model and the kinetics dataset. CoRR, abs/1705.07750, 2017

arXiv 2017
[6]

Chen and C

T. Chen and C. Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining , pages 785–794. ACM, 2016

work page 2016
[7]

K. Choi, G. Fazekas, M. B. Sandler, and K. Cho. Convo- lutional recurrent neural networks for music classiﬁcation. CoRR, abs/1609.04243, 2016

work page Pith review arXiv 2016
[8]

T. Cour, C. Jordan, E. Miltsakaki, and B. Taskar. Movie/script: Alignment and parsing of video and text tran- scription. In European Conference on Computer Vision , pages 158–171. Springer, 2008

work page 2008

Show all 39 references

[9]

Everingham, J

M. Everingham, J. Sivic, and A. Zisserman. Hello! my name is... buffy” – automatic naming of characters in tv video. In BMVC, 2006

2006
[10]

Glorot and Y

X. Glorot and Y . Bengio. Understanding the difﬁculty of training deep feedforward neural networks. In In Proceed- ings of the International Conference on Artiﬁcial Intelligence and Statistics (AISTATS10). Society for Artiﬁcial Intelligence and Statistics, 2010

2010
[11]

F. C. Heilbron, V . Escorcia, B. Ghanem, and J. C. Niebles. Activitynet: A large-scale video benchmark for human activ- ity understanding. In Computer Vision and Pattern Recog- nition (CVPR), 2015 IEEE Conference on , pages 961–970. IEEE, 2015

2015
[12]

Hochreiter and J

S. Hochreiter and J. Schmidhuber. Long short-term memory. Neural Comput., 9(8):1735–1780, Nov. 1997

1997
[13]

Idrees, A

H. Idrees, A. R. Zamir, Y .-G. Jiang, A. Gorban, I. Laptev, R. Sukthankar, and M. Shah. The thumos challenge on ac- tion recognition for videos in the wild. Computer Vision and Image Understanding, 155:1–23, 2017

2017
[14]

Joulin, E

A. Joulin, E. Grave, P. Bojanowski, and T. Mikolov. Bag of tricks for efﬁcient text classiﬁcation. CoRR, abs/1607.01759, 2016

2016 arXiv
[15]

Karpathy, G

A. Karpathy, G. Toderici, S. Shetty, T. Leung, R. Sukthankar, and L. Fei-Fei. Large-scale video classiﬁcation with convo- lutional neural networks. In Proceedings of the IEEE con- ference on Computer Vision and Pattern Recognition, pages 1725–1732, 2014

2014
[16]

Karpathy, G

A. Karpathy, G. Toderici, S. Shetty, T. Leung, R. Sukthankar, and L. Fei-Fei. Large-scale video classiﬁcation with convo- lutional neural networks. In 2014 IEEE Conference on Com- puter Vision and Pattern Recognition , pages 1725–1732, June 2014

2014
[17]

D. P. Kingma and J. Ba. Adam: A method for stochastic optimization. CoRR, abs/1412.6980, 2014

2014 arXiv
[18]

Kuehne, H

H. Kuehne, H. Jhuang, E. Garrote, T. Poggio, and T. Serre. HMDB: a large video database for human motion recog- nition. In Proceedings of the International Conference on Computer Vision (ICCV), 2011

2011
[19]

Mikolov, I

T. Mikolov, I. Sutskever, K. Chen, G. S. Corrado, and J. Dean. Distributed representations of words and phrases and their compositionality. In Advances in neural informa- tion processing systems, pages 3111–3119, 2013

2013
[20]

Pennington, R

J. Pennington, R. Socher, and C. D. Manning. Glove: Global vectors for word representation. In Empirical Methods in Natural Language Processing (EMNLP), pages 1532–1543, 2014

2014
[21]

Ramanathan, A

V . Ramanathan, A. Joulin, P. Liang, and L. Fei-Fei. Linking people in videos with their names using coreference resolu- tion. In European Conference on Computer Vision , pages 95–110. Springer, 2014

2014
[22]

K. K. Reddy and M. Shah. Recognizing 50 human action categories of web videos. Machine Vision and Applications, 24(5):971–981, 2013

2013
[23]

Rohrbach, A

A. Rohrbach, A. Torabi, M. Rohrbach, N. Tandon, C. Pal, H. Larochelle, A. Courville, and B. Schiele. Movie descrip- tion. International Journal of Computer Vision, 123(1):94– 120, 2017

2017
[24]

Sankar, C

P. Sankar, C. Jawahar, and A. Zisserman. Subtitle-free movie to script alignment. In Proc. Brit. Mach. Vis. Conf. , pages 121–1, 2009

2009
[25]

Schuldt, I

C. Schuldt, I. Laptev, and B. Caputo. Recognizing human actions: a local svm approach. In Pattern Recognition, 2004. ICPR 2004. Proceedings of the 17th International Confer- ence on, volume 3, pages 32–36. IEEE, 2004

2004
[26]

G. A. Sigurdsson, G. Varol, X. Wang, A. Farhadi, I. Laptev, and A. Gupta. Hollywood in homes: Crowdsourcing data collection for activity understanding. In European Confer- ence on Computer Vision, pages 510–526. Springer, 2016

2016
[27]

G. S. Simoes, J. Wehrmann, R. C. Barros, and D. D. Ruiz. Movie genre classiﬁcation with convolutional neural net- works. In 2016 International Joint Conference on Neural Networks, IJCNN 2016, Vancouver, BC, Canada, July 24-29, 2016, pages 259–266, 2016

2016
[28]

Simonyan and A

K. Simonyan and A. Zisserman. Two-stream convolu- tional networks for action recognition in videos. CoRR, abs/1406.2199, 2014

2014 arXiv
[29]

Simonyan and A

K. Simonyan and A. Zisserman. Very deep convolu- tional networks for large-scale image recognition. CoRR, abs/1409.1556, 2014

2014 arXiv
[30]

Sivic, M

J. Sivic, M. Everingham, and A. Zisserman. who are you?- learning person speciﬁc classiﬁers from video. In Computer Vision and Pattern Recognition, 2009. CVPR 2009. IEEE Conference on, pages 1145–1152. IEEE, 2009

2009
[31]

M. Slaney. Web-scale multimedia analysis: Does content matter? IEEE MultiMedia, 18(2):12–15, 2011

2011
[32]

Soomro, A

K. Soomro, A. R. Zamir, and M. Shah. UCF101: A dataset of 101 human actions classes from videos in the wild.CoRR, abs/1212.0402, 2012

2012 arXiv
[33]

Srivastava, G

N. Srivastava, G. Hinton, A. Krizhevsky, I. Sutskever, and R. Salakhutdinov. Dropout: A simple way to prevent neu- ral networks from overﬁtting. Journal of Machine Learning Research, 15:1929–1958, 2014

1929
[34]

Srivastava, E

N. Srivastava, E. Mansimov, and R. Salakhutdinov. Unsuper- vised learning of video representations using lstms. CoRR, abs/1502.04681, 2015

2015 arXiv
[35]

Tapaswi, M

M. Tapaswi, M. Bauml, and R. Stiefelhagen. Book2movie: Aligning video scenes with book chapters. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 1827–1835, 2015

2015
[36]

Tapaswi, Y

M. Tapaswi, Y . Zhu, R. Stiefelhagen, A. Torralba, R. Ur- tasun, and S. Fidler. Movieqa: Understanding stories in movies through question-answering. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recogni- tion, pages 4631–4640, 2016

2016
[37]

D. Tran, L. D. Bourdev, R. Fergus, L. Torresani, and M. Paluri. C3D: generic features for video analysis. CoRR, abs/1412.0767, 2014

2014 arXiv
[38]

H. Zhou, T. Hermans, A. V . Karandikar, and J. M. Rehg. Movie genre classiﬁcation via scene categorization. In Pro- ceedings of the 18th ACM International Conference on Mul- timedia, MM ’10, pages 747–750, New York, NY , USA,
[39]

Y . Zhu, R. Kiros, R. Zemel, R. Salakhutdinov, R. Urtasun, A. Torralba, and S. Fidler. Aligning books and movies: To- wards story-like visual explanations by watching movies and reading books. In Proceedings of the IEEE international conference on computer vision, pages 19–27, 2015

2015

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Abu-El-Haija, N

S. Abu-El-Haija, N. Kothari, J. Lee, P. Natsev, G. Toderici, B. Varadarajan, and S. Vijayanarasimhan. Youtube-8m: A large-scale video classiﬁcation benchmark. CoRR, abs/1609.08675, 2016

arXiv 2016

[2] [2]

Bamman, B

D. Bamman, B. O’Connor, and N. A. Smith. Learning la- tent personas of ﬁlm characters. In Proceedings of the An- nual Meeting of the Association for Computational Linguis- tics (ACL), page 352, 2014

work page 2014

[3] [3]

B ¨auml, M

M. B ¨auml, M. Tapaswi, and R. Stiefelhagen. Semi- supervised learning with constraints for person identiﬁcation in multimedia data. In Computer Vision and Pattern Recog- nition (CVPR), 2013 IEEE Conference on, pages 3602–3609. IEEE, 2013

work page 2013

[4] [4]

Bojanowski, F

P. Bojanowski, F. Bach, I. Laptev, J. Ponce, C. Schmid, and J. Sivic. Finding actors and actions in movies. In Com- puter Vision (ICCV), 2013 IEEE International Conference on, pages 2280–2287. IEEE, 2013

work page 2013

[5] [5]

Carreira and A

J. Carreira and A. Zisserman. Quo vadis, action recog- nition? A new model and the kinetics dataset. CoRR, abs/1705.07750, 2017

arXiv 2017

[6] [6]

Chen and C

T. Chen and C. Guestrin. Xgboost: A scalable tree boosting system. In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining , pages 785–794. ACM, 2016

work page 2016

[7] [7]

K. Choi, G. Fazekas, M. B. Sandler, and K. Cho. Convo- lutional recurrent neural networks for music classiﬁcation. CoRR, abs/1609.04243, 2016

work page Pith review arXiv 2016

[8] [8]

T. Cour, C. Jordan, E. Miltsakaki, and B. Taskar. Movie/script: Alignment and parsing of video and text tran- scription. In European Conference on Computer Vision , pages 158–171. Springer, 2008

work page 2008