WSLLN: Weakly Supervised Natural Language Localization Networks

Caiming Xiong; Larry S. Davis; Mingfei Gao; Richard Socher

REVIEW 4 major objections 6 minor 1 cited by

WSLLN: Weakly Supervised Natural Language Localization Networks

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 A two-branch network localizes video moments using only video-sentence pairs.

desk verdict A real but modest weakly-supervised moment-localization result; the pseudo-label refinement is the load-bearing part and it gets the least analysis. read the letter →

arxiv 1909.00239 v1 pith:HSIJ7IUA submitted 2019-08-31 cs.CV

Mingfei Gao , Larry S. Davis , Richard Socher , Caiming Xiong This is my paper · ORCID

classification cs.CV

keywords weaklysupervisedlearningnaturallanguagelocalizationtemporalgroundingvideo-sentencepairspseudolabelstwo-branchnetworkActivityNetCaptionsDiDeMo

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

The paper sets out to show that natural language localization in untrimmed video—finding the moment a sentence describes—does not require start and end time annotations for training. It proposes WSLLN, a two-branch network trained end-to-end on video-sentence pairs alone: one branch scores how well each video segment matches the query, the other selects among segments, and the two scores are multiplied and summed into a video-level matching score. A pseudo-label refinement term, generated from the network's own best-scoring segment, pushes training beyond coarse video-level matching. On ActivityNet Captions the method reports a weakly supervised mean IoU of 32.2, about 4 points above the previous weak baseline WSDEC-W and above the supervised CTRL; on DiDeMo it gives the first reported weakly supervised results, behind the supervised MCN. If correct, the paper shows that the expensive segment-level supervision used by most moment-localization models can be replaced by cheap video-sentence correspondence.

What carries the argument

The load-bearing mechanism is the two-branch score fusion with a pseudo-label refinement loop. Alignment branch computes per-proposal consistency $s^a_i = \operatorname{softmax}_a(W_a f^m_i)$; detection branch computes competition across proposals $s^d_i = \operatorname{softmax}_d(W_d f^m_i)$; the merged score is $s_i = s^a_i \cdot s^d_i$, and the video-level match score is $vq_i = \sum_j s^j_i$. Training minimizes $L = \operatorname{loss}(vq_i, l_i) + \lambda \operatorname{loss}(s^j_i, \hat y_i)$, where $\hat y_i = \arg\max_j s^j_i[:,1]$ is the pseudo segment label generated by the model itself. This machinery is what converts video-level labels into per-segment supervision, and the pseudo-label term is what lets the model refine segment selection without ever seeing human start and end times.

What would settle it

On a held-out subset of ActivityNet Captions where human start and end times exist but are withheld from training, compute the IoU between the network's argmax pseudo label and the human segment during training; if the pseudo labels match the human segment no better than randomly chosen proposals, yet the refinement branch still lifts mIoU, then the refinement term is not learning genuine segment-text correspondence and the claimed weak supervision gain would not transfer to a dataset where the initial guesses are biased.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is that the temporal segment matching a sentence can be learned from weak video-sentence labels by decomposing the task into two complementary decisions. The alignment branch independently scores each proposal's consistency with the query, while the detection branch applies a softmax across proposals so that segments compete; their elementwise product is the final proposal score. Summing these scores over proposals turns the problem into video-sentence matching, trainable with positive pairs and randomly constructed negative pairs. An auxiliary loss uses the argmax proposal as a pseudo segment label, adding a refinement signal without human coordinates. The result is a single-stage, end-to-end model that reaches mIoU 32.2 on ActivityNet Captions in the weakly supervised setting, surpassing WSDEC-W by roughly 4 mIoU and the supervised CTRL by more than 11 mIoU, while remaining below stronger supervised models such as ABLR and WSDEC-S.

Load-bearing premise

The load-bearing premise is that the segment the network itself scores highest during training is a trustworthy enough label to teach it where the event is; if that self-generated pseudo label is wrong, the refinement loss will reinforce the model's existing mistakes instead of fixing them.

Editorial extensions

If this is right

If the central claim holds, moment localization can be trained from any corpus of paired videos and sentences—subtitles, narration, or text-based video retrieval output—cutting the main annotation cost of temporal grounding.
The two-branch design shows that proposal competition plus video-level matching is sufficient to induce segment selection, extending a recipe from weakly supervised object detection to video grounding.
The pseudo-label refinement term provides a consistent gain when $\lambda$ is between 0.1 and 0.4, so the method is not critically sensitive to the balancing weight in that range.
On smaller datasets such as DiDeMo, the gap to supervised models widens, implying that weakly supervised localization benefits from more training video-sentence pairs; scaling data may matter more than architectural complexity.
Swapping BERT for GRU costs only about 0.4 mIoU, so the reported strong result is not driven by the sentence encoder; the training formulation itself carries the improvement.

Reading between the lines

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

Editorial inference: because only pair-level labels are needed, automatically generated captions or narrated video could be used to scale training data far beyond manually annotated grounding datasets; the paper does not test this but it follows directly from the weak supervision setup.
Editorial inference: the pseudo-label refinement is self-training, so on videos where the true moment falls outside all proposals, the argmax pseudo label will reinforce the model's blind spot. A simple test would be to withhold the top proposal and use the second-best segment as the pseudo label and compare mIoU.
Editorial inference: the selection branch's softmax over proposals is a soft attention mechanism, so the same loss could be adapted to other weakly supervised video-language tasks such as paragraph grounding or video question answering with only video-answer pairs.
Editorial inference: the paper's headline comparison uses different sentence encoders (BERT vs GRU in WSDEC-W); although the GRU ablation suggests the encoder is not the decisive factor, a fully matched head-to-head with WSDEC-W's exact training procedure would isolate the contribution of the two-branch design.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Colleague,

The paper delivers a usable weakly supervised method for natural language moment localization on untrimmed video. The headline numbers are real: on ActivityNet Captions, WSLLN gets 32.2 mIoU against WSDEC-W's 28.2, and it's the first weak-supervision result reported on DiDeMo. The gains over prior weak work are modest but consistent, and the gap to supervised methods is acknowledged openly.

What's good: the architecture is simple and end-to-end, a clean translation of WSDDN's two-branch idea to language grounding. The GRU-for-BERT swap shows the encoder choice isn't the main driver of performance. The two-branch ablation shows both terms contribute substantially, though the "Align-only" and "Detect-only" baselines are constructed a bit crudely.

The soft spots, in order of importance. First, the refinement loss in Eq. 6 is self-referential: the pseudo-label is the argmax of the same scores being trained. The paper doesn't analyze how often those pseudo-labels are correct, whether they stabilize, or whether the 4.8-point gain from the refinement branch comes from fixing alignment errors or reinforcing a bias (like always picking an early or long proposal). That's a legitimate concern that the authors need to address—a simple experiment with ground-truth proposals or decoupled labels would clarify. Second, lambda is selected on the same validation split where final numbers are reported; the flat curve in Table 2 makes this less damaging, but it's still a risk. Third, no code and minimal training details are given, so a full reproduction isn't possible from the paper. Fourth, the abstract's "state-of-the-art" is too broad: Table 1 shows WSDEC-W is still slightly ahead at IoU=0.5, and supervised methods are ahead across the board. The claim should be properly qualified.

None of this undermines the central point—that video-sentence pairs alone can train a moment localizer with useful accuracy. It just means the result is thinner than the abstract suggests, and the refinement branch needs a more careful supporting analysis.

This paper deserves peer review, with requests for a pseudo-label quality study, full training details, and a toned-down SOTA claim. I'd bring it to a reading group to unpack the self-training loop.

Referee Report

4 major / 6 minor

Summary. The paper proposes WSLLN, a weakly supervised network for natural language moment localization in untrimmed videos. Training uses only video-sentence pairs, with no temporal annotations. The model extracts proposal and sentence features, fuses them, and passes them through two branches: an alignment branch that scores each proposal's semantic match with the text, and a detection branch that competitively selects proposals. The two score sets are merged by element-wise multiplication, and training uses a video-level matching loss plus an auxiliary refinement loss whose pseudo labels are taken from the network's own argmax over proposals. Experiments on ActivityNet Captions and DiDeMo report weakly supervised mIoU of 32.2 on ActivityNet Captions, an ablation of the refinement weight, comparison to a GRU sentence encoder, and two-branch ablations.

Significance. If the central claim holds, the contribution is useful: WSLLN is a simple end-to-end model that uses only video-sentence pairs, shows a substantial mIoU gain over the previously reported weakly supervised WSDEC-W on ActivityNet Captions, and includes a GRU variant indicating that the gain is not solely due to BERT. The paper is honest in reporting that supervised baselines still perform better, and the three-run variance is small. However, the headline gain is currently entangled with a self-supervised pseudo-label loop, and the comparison to prior work is limited to one weakly supervised baseline on ActivityNet and no weakly supervised baseline on DiDeMo. The significance is therefore conditional on additional evidence that the refinement branch corrects rather than reinforces the model's biases.

major comments (4)

[§3.2, Eq. (6); §4.2.2, Table 2] The pseudo-label refinement loss is self-referential and its measured benefit is not yet separated from self-supervision bias. The label is defined as yhat_i = argmax_j s_j^i[:,1], i.e., the argmax over the very scores being optimized, so L_r trains the model to concentrate probability mass on its own current top proposal. Because the video-level loss L_v sums scores over all proposals, it does not determine which proposal is selected; the refinement term is what shapes the final segment choice. Table 2 shows that removing this term (lambda=0) reduces mIoU from 32.2 to 27.4, so the claimed 4.8-point improvement rests entirely on this loop. The paper reports no analysis of pseudo-label accuracy against ground truth, no information on how pseudo labels evolve during training, and no experiment that decouples the label source from the trained model. The concern is not merely theoretical: the improvement could reflect reinforcement of a length or position prior rather than correction of alignment errors. I request three concrete tests: (i) report the agreement rate between yhat_i and ground-truth segments on the validation set over training; (ii) replace yhat_i with fixed pseudo labels generated by a separately trained model (or with ground-truth labels in an oracle ablation) and measure whether the gain persists; and (iii) report the lambda selection procedure explicitly, since Table 2 evaluates lambda on the same validation split used for the headline numbers, which conflates model selection with architecture effectiveness.
[§4.2.1, Table 1] The headline claim that WSLLN 'largely outperforms' WSDEC-W by about 4 mIoU is not uniformly supported across thresholds. WSLLN is higher at IoU=0.1 (75.4 vs 62.7) and marginally higher at IoU=0.3 (42.8 vs 42.0), but lower at IoU=0.5 (22.7 vs 23.3). The paper should explicitly discuss this threshold-dependent behavior and temper the claim. A method with higher mIoU but lower strict-IoU recall is not unambiguously better for practical localization, and the discrepancy also suggests that the pseudo-label refinement may be biasing predictions toward partially overlapping but not precisely localized segments.
[§4.1, §4.2.1, §4.3] The experimental report lacks the training details needed to verify the central empirical claim. No optimizer, learning rate, batch size, number of epochs, negative-sampling procedure (beyond 'randomly selected sentence'), or pseudo-label update schedule is given, and no code is released. These details matter because the method is a training-time mechanism: the refinement loss, the balancing of L_v and L_r, and the generation of negatives all affect the reported numbers. Without them or a public implementation, the comparisons in Tables 1 and 4 cannot be reproduced. I also note that Table 4 contains no weakly supervised baseline, so the claim of 'state-of-the-art' is only supported on ActivityNet Captions and only relative to WSDEC-W.
[§3.2, Eq. (7), Eq. (6)] The formulation of the merged scores and the refinement loss is ambiguous. The text says si = sai · sdi, but both sai and sdi are n x 2 matrices, so it is unclear whether the operation is element-wise multiplication, matrix multiplication, or a different fusion; Figure 1 labels 'dot prod' but the axes are not specified. Relatedly, Eq. (6) writes loss(sj_i, yhat_i) without stating whether sj_i is the merged score, the alignment-branch score, or the detection-branch score used for pseudo-label generation. Since the pseudo-label is defined as argmax_j s_j^i[:,1], the object whose column 1 is used must be stated precisely. This ambiguity affects both the correctness of the derivation and the reproducibility of Table 2.

minor comments (6)

[Table 4, §4.3] The text 'Following MCN, we set th = 1.0 for the IoU threshold' is almost certainly a typo; with IoU threshold 1.0, R@1 would require exact segment matches and the reported values above 15% would be implausible. Please state the intended threshold (e.g., 0.5) and why it is used for DiDeMo.
[Figure 1] The figure caption is garbled (e.g., 'input sentence is here', '×...#$:&×'×''), and the axis labels 'x×z indicates dimensions' are unclear. The caption should be rewritten to describe the tensor shapes and the fusion operation legibly.
[Eq. (1), Eq. (2)] The notation for normalized proposal times is underspecified: '¯j means time is normalized to [0, 1]' should be written as separate start and end values, and the concatenation in Eq. (1) should specify which tensors are concatenated along which axis. In Eq. (2), the use of FC(fp||fq) inside the multimodal feature should be defined, including the output dimension and nonlinearity.
[Table 3] Table 3 would be easier to read if the full WSLLN row were included alongside Align-only and Detect-only, so that the reader can see the gains from the complete architecture in a single table.
[§4.2.1] The comparison to CTRL should be contextualized: CTRL uses different proposal features and a different training objective, so the 'over 11% mIoU' improvement over CTRL is not an apples-to-apples architectural comparison. The sentence should note the differences or restrict the claim to the weakly supervised setting.
[Abstract and Conclusion] The phrase 'state-of-the-art performance' in the abstract is too strong given that supervised methods outperform WSLLN on both datasets and that the weakly supervised comparison set is small. Please qualify it as state-of-the-art among weakly supervised methods on ActivityNet Captions.

Circularity Check

1 steps flagged · score 5.0 of 10

Pseudo-label refinement is self-referential, but the external video-sentence loss and held-out evaluation keep the main claim partially independent.

self definitional [Section 3.2, Training Phase, Eq. 6]
"Results can be further refined by adding an auxiliary task L_r in Eq. 6 where ŷ_i = {0,1,...,n−1} indicates the index of the segment that best matches the sentence during training. The real segment-level labels are not available, thus we generate pseudo labels by setting ŷ_i = argmax_j s_j^i[:,1]. ... L_r = loss(s_j^i, ŷ_i)."

The pseudo-label ŷ_i is defined as the argmax over the very scores s_j^i that L_r optimizes. With one-hot cross-entropy, minimizing L_r increases s_{ŷ_i} and suppresses the other proposals, so the auxiliary objective is exactly a self-agreement constraint: the model is trained to concentrate mass on its own current top proposal. The video-level loss L_v only supervises the sum Σ_j s_j^i, so it does not determine which individual proposal is selected; the argmax choice is therefore shaped by this self-referential pseudo-label rather than by independent segment-level evidence. Consequently, the Table 2 improvement when λ goes from 0 to 0.1 is not by itself evidence that the selected segment is more correct; it can reflect the model entrenching its existing bias.

full rationale

The paper's core weakly supervised formulation is not circular in itself: training uses only video-sentence pairs, comparison baselines are external, and the headline mIoU numbers are measured against ground-truth temporal segments on held-out splits. The two-branch design and the use of pseudo-labels are also standard weakly supervised techniques, not disguised fits of the final evaluation metric. The one structurally self-referential component is the auxiliary refinement loss in Eq. 6: its label ŷ_i is the argmax of the same scores being trained, so the refinement branch is literally training the model to agree with its own current selection. Since L_v constrains only the sum over proposals, the pseudo-label term is what determines which proposal is ultimately chosen, making the selection partly a self-fulfilling loop. However, the loop is anchored by L_v and by held-out evaluation, so the central claim does not reduce entirely to a fit or a self-citation chain. A score of 5 reflects one load-bearing self-referential training step with real but partial external mitigation.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the MIL assumption that video-level sentence matching trains segment-level alignment, on a self-generated pseudo-label loop whose contribution is demonstrated only by ablation, and on frozen features and fixed proposal sets from prior work. No new physical or conceptual entities are introduced.

free parameters (1)

lambda (Eq. 7) = 0.2 (stable range 0.1-0.4)
Balancing weight between video-sentence matching loss L_v and pseudo-segment refinement loss L_r; selected by the validation sweep in Table 2 and directly affects reported scores.

assumptions (4)

domain assumption Aggregate video-sentence matching score, v_q = sum_j s_j, is a sufficient training signal to learn per-proposal alignment (Eq. 5).
The multiple instance learning assumption that at least one proposal matches the query and can be supervised by video-level labels is unproved and is the foundation of the weak supervision.
ad hoc to paper Pseudo labels from the model's own argmax, yhat_i = argmax_j s_j[:,1], are reliable enough to improve localization via L_r (Eq. 6).
Self-generated labels are a modeling choice specific to this paper; Table 2 shows performance depends on this auxiliary loss.
domain assumption Pretrained C3D/VGG video features and BERT text features retain enough semantic and temporal information for localization (Sec 4.1).
The method does not fine-tune visual or language encoders; all localization capacity is assumed to be present in the frozen features.
domain assumption Provided proposal sets (15 from Duan et al, 21 from Hendricks et al) cover the ground-truth events well enough for ranking and mIoU evaluation (Sec 4.1).
Localization is restricted to these proposals; poor coverage would cap achievable mIoU regardless of the learned scores.

how reviews work

0 comments

Cite this review

Pith. "Pith review of WSLLN: Weakly Supervised Natural Language Localization Networks." pith.science (2026). https://pith.science/paper/HSIJ7IUA

@misc{pith2026190900239,
  author       = {Pith},
  title        = {Pith review of: WSLLN: Weakly Supervised Natural Language Localization Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HSIJ7IUA}},
  note         = {Machine review of arXiv:1909.00239}
}

read the original abstract

We propose weakly supervised language localization networks (WSLLN) to detect events in long, untrimmed videos given language queries. To learn the correspondence between visual segments and texts, most previous methods require temporal coordinates (start and end times) of events for training, which leads to high costs of annotation. WSLLN relieves the annotation burden by training with only video-sentence pairs without accessing to temporal locations of events. With a simple end-to-end structure, WSLLN measures segment-text consistency and conducts segment selection (conditioned on the text) simultaneously. Results from both are merged and optimized as a video-sentence matching problem. Experiments on ActivityNet Captions and DiDeMo demonstrate that WSLLN achieves state-of-the-art performance.

Figures

Figures reproduced from arXiv: 1909.00239 by the authors.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Weakly Supervised Temporal Sentence Grounding via Positive Sample Mining
cs.CV 2025-05 conditional novelty 6.0 of 10

A positive sample mining method that groups training queries by text similarity and uses contrastive plus rank losses improves weakly supervised temporal sentence grounding on Charades-STA, ActivityNet Captions, and NExT-GQA.

Reference graph

Works this paper leans on

42 extracted references · 35 canonical work pages · cited by 1 Pith paper

[1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...
[3]

Hakan Bilen and Andrea Vedaldi. 2016. Weakly supervised deep detection networks. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

work page 2016
[4]

Piotr Bojanowski, R \'e mi Lajugie, Francis Bach, Ivan Laptev, Jean Ponce, Cordelia Schmid, and Josef Sivic. 2014. Weakly supervised action labeling in videos under ordering constraints. In European Conference on Computer Vision, pages 628--643. Springer

work page 2014
[5]

Shyamal Buch, Victor Escorcia, Chuanqi Shen, Bernard Ghanem, and Juan Carlos Niebles. 2017. SST : Single-stream temporal action proposals. In CVPR

work page 2017
[6]

Yu-Wei Chao, Sudheendra Vijayanarasimhan, Bryan Seybold, David A Ross, Jia Deng, and Rahul Sukthankar. 2018. Rethinking the faster r-cnn architecture for temporal action localization. In CVPR

work page 2018
[7]

Kyunghyun Cho, Bart Van Merri \"e nboer, Caglar Gulcehre, Dzmitry Bahdanau, Fethi Bougares, Holger Schwenk, and Yoshua Bengio. 2014. Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv:1406.1078

arXiv 2014
[8]

Xiyang Dai, Bharat Singh, Guyue Zhang, Larry S Davis, and Yan Qiu Chen. 2017. Temporal context network for activity localization in videos. In ICCV

work page 2017

Show all 42 references

[9]

Roeland De Geest, Efstratios Gavves, Amir Ghodrati, Zhenyang Li, Cees Snoek, and Tinne Tuytelaars. 2016. Online action detection. In ECCV

2016
[10]

Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805

2018 arXiv
[11]

Ali Diba, Vivek Sharma, Ali Pazandeh, Hamed Pirsiavash, and Luc Van Gool. 2017. Weakly supervised cascaded convolutional networks. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pages 5131--5139

2017
[12]

Xuguang Duan, Wenbing Huang, Chuang Gan, Jingdong Wang, Wenwu Zhu, and Junzhou Huang. 2018. Weakly supervised dense event captioning in videos. In Advances in Neural Information Processing Systems, pages 3059--3069

2018
[13]

Olivier Duchenne, Ivan Laptev, Josef Sivic, Francis R Bach, and Jean Ponce. 2009. Automatic annotation of human actions in video. In ICCV, volume 1, pages 3--2

2009
[14]

Jiyang Gao, Chen Sun, Zhenheng Yang, and Ram Nevatia. 2017 a . Tall: Temporal activity localization via language query. In Proceedings of the IEEE International Conference on Computer Vision, pages 5267--5275

2017
[15]

Jiyang Gao, Zhenheng Yang, and Ram Nevatia. 2017 b . RED : Reinforced encoder-decoder networks for action anticipation. In BMVC

2017
[16]

Jiyang Gao, Zhenheng Yang, Chen Sun, Kan Chen, and Ram Nevatia. 2017 c . TURN TAP : Temporal unit regression network for temporal action proposals. ICCV

2017
[17]

Mingfei Gao, Ang Li, Ruichi Yu, Vlad I Morariu, and Larry S Davis. 2018. C-wsl: Count-guided weakly supervised localization. In ECCV

2018
[18]

Mingfei Gao, Mingze Xu, Larry S Davis, Richard Socher, and Caiming Xiong. 2019. Startnet: Online detection of action start in untrimmed videos. arXiv preprint arXiv:1903.09868

2019 arXiv
[19]

Lisa Anne Hendricks, Oliver Wang, Eli Shechtman, Josef Sivic, Trevor Darrell, and Bryan Russell. 2017. Localizing moments in video with natural language. In Proceedings of the IEEE International Conference on Computer Vision, pages 5803--5812

2017
[20]

Lisa Anne Hendricks, Oliver Wang, Eli Shechtman, Josef Sivic, Trevor Darrell, and Bryan Russell. 2018. Localizing moments in video with temporal language. In Empirical Methods in Natural Language Processing (EMNLP)

2018
[21]

Ronghang Hu, Huazhe Xu, Marcus Rohrbach, Jiashi Feng, Kate Saenko, and Trevor Darrell. 2016. Natural language object retrieval. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 4555--4564

2016
[22]

De-An Huang, Li Fei-Fei, and Juan Carlos Niebles. 2016. Connectionist temporal modeling for weakly supervised action labeling. In European Conference on Computer Vision, pages 137--153. Springer

2016
[23]

Zequn Jie, Yunchao Wei, Xiaojie Jin, Jiashi Feng, and Wei Liu. 2017. Deep self-taught learning for weakly supervised object localization. IEEE CVPR

2017
[24]

Vadim Kantorov, Maxime Oquab, Minsu Cho, and Ivan Laptev. 2016. Contextlocnet: Context-aware deep network models for weakly supervised localization. In European Conference on Computer Vision, pages 350--365. Springer

2016
[25]

Thomas N Kipf and Max Welling. 2016. Semi-supervised classification with graph convolutional networks. arXiv preprint arXiv:1609.02907

2016 arXiv
[26]

Ranjay Krishna, Kenji Hata, Frederic Ren, Li Fei-Fei, and Juan Carlos Niebles. 2017. Dense-captioning events in videos. In International Conference on Computer Vision (ICCV)

2017
[27]

Ivan Laptev, Marcin Marszalek, Cordelia Schmid, and Benjamin Rozenfeld. 2008. Learning realistic human actions from movies. In CVPR

2008
[28]

Dong Li, Jia-Bin Huang, Yali Li, Shengjin Wang, and Ming-Hsuan Yang. 2016. Weakly supervised object localization with progressive domain adaptation. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

2016
[29]

Bingbin Liu, Serena Yeung, Edward Chou, De-An Huang, Li Fei-Fei, and Juan Carlos Niebles. 2018. Temporal modular networks for retrieving complex compositional activities in videos. In Proceedings of the European Conference on Computer Vision (ECCV), pages 552--568

2018
[30]

Maxime Oquab, L \'e on Bottou, Ivan Laptev, and Josef Sivic. 2015. Is object localization for free?-weakly-supervised learning with convolutional neural networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 685--694

2015
[31]

Dim P Papadopoulos, Jasper RR Uijlings, Frank Keller, and Vittorio Ferrari. 2017. Training object class detectors with click supervision. CVPR

2017
[32]

Zheng Shou, Hang Gao, Lei Zhang, Kazuyuki Miyazawa, and Shih-Fu Chang. 2018 a . Autoloc: Weakly-supervised temporal action localization in untrimmed videos. In Proceedings of the European Conference on Computer Vision (ECCV), pages 154--171

2018
[33]

Zheng Shou, Junting Pan, Jonathan Chan, Kazuyuki Miyazawa, Hassan Mansour, Anthony Vetro, Xavier Giro-i Nieto, and Shih-Fu Chang. 2018 b . Online action detection in untrimmed, streaming videos-modeling and evaluation. In ECCV

2018
[34]

Zheng Shou, Dongang Wang, and Shih-Fu Chang. 2016. Temporal action localization in untrimmed videos via multi-stage cnns. In CVPR

2016
[35]

Karen Simonyan and Andrew Zisserman. 2014. Very deep convolutional networks for large-scale image recognition. arXiv:1409.1556

2014 arXiv
[36]

Peng Tang, Xinggang Wang, Xiang Bai, and Wenyu Liu. 2017. Multiple instance detection network with online instance classifier refinement. CVPR

2017
[37]

Du Tran, Lubomir Bourdev, Rob Fergus, Lorenzo Torresani, and Manohar Paluri. 2015. Learning spatiotemporal features with 3d convolutional networks. In ICCV

2015
[38]

Limin Wang, Yuanjun Xiong, Dahua Lin, and Luc Van Gool. 2017. Untrimmednets for weakly supervised action recognition and detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pages 4325--4334

2017
[39]

Mingze Xu, Mingfei Gao, Yi-Ting Chen, Larry S Davis, and David J Crandall. 2018. Temporal recurrent networks for online action detection. arXiv:1811.07391

2018 arXiv
[40]

Yitian Yuan, Tao Mei, and Wenwu Zhu. 2018. To find where you talk: Temporal sentence localization in video with attention based location regression. arXiv preprint arXiv:1804.07014

2018 arXiv
[41]

Da Zhang, Xiyang Dai, Xin Wang, Yuan-Fang Wang, and Larry S Davis. 2018. Man: Moment alignment network for natural language moment retrieval via iterative graph adjustment. arXiv preprint arXiv:1812.00087

2018 arXiv
[42]

Yue Zhao, Yuanjun Xiong, Limin Wang, Zhirong Wu, Xiaoou Tang, and Dahua Lin. 2017. Temporal action detection with structured segment networks. In ICCV

2017

Pith tools

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

[1] [1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

[3] [3]

Hakan Bilen and Andrea Vedaldi. 2016. Weakly supervised deep detection networks. In The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)

work page 2016

[4] [4]

Piotr Bojanowski, R \'e mi Lajugie, Francis Bach, Ivan Laptev, Jean Ponce, Cordelia Schmid, and Josef Sivic. 2014. Weakly supervised action labeling in videos under ordering constraints. In European Conference on Computer Vision, pages 628--643. Springer

work page 2014

[5] [5]

Shyamal Buch, Victor Escorcia, Chuanqi Shen, Bernard Ghanem, and Juan Carlos Niebles. 2017. SST : Single-stream temporal action proposals. In CVPR

work page 2017

[6] [6]

Yu-Wei Chao, Sudheendra Vijayanarasimhan, Bryan Seybold, David A Ross, Jia Deng, and Rahul Sukthankar. 2018. Rethinking the faster r-cnn architecture for temporal action localization. In CVPR

work page 2018

[7] [7]

Kyunghyun Cho, Bart Van Merri \"e nboer, Caglar Gulcehre, Dzmitry Bahdanau, Fethi Bougares, Holger Schwenk, and Yoshua Bengio. 2014. Learning phrase representations using rnn encoder-decoder for statistical machine translation. arXiv:1406.1078

arXiv 2014

[8] [8]

Xiyang Dai, Bharat Singh, Guyue Zhang, Larry S Davis, and Yan Qiu Chen. 2017. Temporal context network for activity localization in videos. In ICCV

work page 2017