ChaLearn Looking at People: Identity-preserved Human Detection (Track 2: Thermal)

Organized by aclapes - Current server time: Dec. 5, 2019, 5 p.m. UTC

Current

Learning
Nov. 19, 2019, midnight UTC

Next

Final Evaluation
Jan. 25, 2020, midnight UTC

End

Competition Ends
April 2, 2020, midnight UTC

 [22/11/2019] ***IMPORTANT: New fixed version of thermal train/valid data (v2)***

 

Short summary of the competition

This is a ChaLearn Looking at People's competition organized in the context of the 15th IEEE International Conference on Automatic Face and Gesture Recognition, Buenos Aires, Argentina, 2020. The mail goal is to develop computer vision methods for human detection in depth and/or thermal images (see Figure 1).

In order to keep duplicate information to the minimum all the information has been centralized in the ChaLearn webpage. For more details, please refer to the following sites:

Challenge description:

http://chalearnlap.cvc.uab.es/challenge/34/description/

IPHD dataset:

http://chalearnlap.cvc.uab.es/dataset/34/description/

 


Figure 1: IPHD dataset sample. First & third columns are depth frames. Second & fourth are their corresponding (aligned-to-depth) thermal frames.

 

Goal of the track

For this competition track, we ask the participants to perform human detection in the thermal modality. Thermal cameras provide temperature readings from the scene. They are less noisy than depth cameras, but at a comparable price they offer a much lower image resolution.

Given the provided thermal images (and bounding box groundtruth annotations), the participants will be asked to develop their thermal-based human detection method. The method will need to output a list of bounding boxes (along with associated confidence scores) per frame containing each person in it. The performance the image-based human detection methods will be evaluated in terms of average precision.

 

Associated events

Top ranking participants will be invited to submit their papers to FG2020 Chalearn LAP Workshop on Privacy-aware Computer Vision.

Evaluation metric

The metric established to evaluate participants' submitted predictions and rank them is Average Precision at IoU=0.50, i.e. AP@0.50, computed from precision-recall curves. For more details on the metrics refer to the explanation provided in the IPHD dataset. The implementation of the metric, the same we use for evaluation, can be found in the starting kit in section "Participate > Data and Starting Kit".

Despite the leaderboard rankings will be based on AP@0.5, we also compute AP@0.25 and AP@0.75 for purposes of clarity. Hopefully, the participants can gather some useful information from those.

 

Evaluation protocol

Participants will need to submit a zip containing one and only file Python pickle file named predictions.pkl. The pickle file will only contain a Python dict where each keys is an image filenames (without file extension) and the value is a list of detections for that image. A detection will be formatted as a tuple of this form: (x, y, w, h, conf), where x and y are the center of the predicted box along the horizontal and the vertical axis respectively, w and h the width and height of the box, and conf the confidence of the prediction. As in the groundtruth annotation files in the Labels/ data folder, the variables x, y, w, h will be normalized in the range [0,1]. Frames with no detections be indicated using an empty list in order to match the number of predicted images and the number of groundtruth images.

As an example, we provide the predictions.pkl submitted as baseline for the competition track in the starting kit in section "Participate > Data and Starting Kit".

 

Baseline submission

For this track, as well as for the depth track, the baseline model is based on the YOLOv3 framework [1]. More precisely, the tiny version of the architecture [2] illustrated in Figure 1.

 

Figure 1. YOLOv3 tiny architecture used in for the baseline model.

 

 

References

[1] https://pjreddie.com/media/files/papers/YOLOv3.pdf

[2] https://github.com/pjreddie/darknet/blob/master/cfg/yolov3-tiny.cfg.

None

CHALEARN Contest Rules for ChaLearn Looking at People

Identity-preserved Human Detection (IPHD) 2020

Official rules

Common terms used in these rules:

These are the official rules that govern how the 2020 ChaLearn Looking at People Challenge Identity-preserved Human Detection contest promotion will operate. This promotion will besimply referred to as the contest or the challenge throughout the rest of these rules and maybe abbreviated on our website, in our documentation, and other publications as ChaLearnLAP 2020.

In these rules, organizers, we, our, and us refer to CHALEARN and "participant”, “you,” and “yourself” refer to an eligible contest participant.

  1. Contest description

    This is a skill-based contest and chance plays no part in the determination of the winner(s). There are three (3) tracks associated to this contest as described below:

    • Depth track. Given the provided depth frames (and bounding box groundtruth annotations), the participants will be asked to develop their depth-based human detection method. Depth cameras are cost-effective devices that provide geometric information of the scene at a resolution and frame acquisition speed that is comparable to RGB cameras. The downside is their noisiness at large real distances. The method developed by the participants will need to output a list of bounding boxes (along with their associated confidences scores) per frame containing each person in it. The performance on depth image-based human detection will be evaluated.

    • Thermal track. Given the provided thermal frames (and bounding box groundtruth annotations), the participants will be asked to develop their thermal-based human detection method. Thermal cameras provide temperature readings from the scene. They are less noisy than depth cameras, but at a comparable price they offer a much lower image resolution. The method developed by the participants will need to output a list of bounding boxes (along with their associated confidences scores) per frame containing each person in it. The performance on depth image-based human detection will be evaluated.

    • Depth-Thermal Fusion track. Given the provided aligned depth-thermal frames (and bounding box groundtruth annotations), the participants will be asked to develop their multimodal (depth and thermal) human detection method. Both modalities have been temporally and spatially aligned and, hence, so they will try to exploit their potential complementarity with a proper fusion strategy. The participants will need to output a list of bounding boxes (along with their associated confidences scores) per frame containing each person in it. The performance on depth image-based human detection will be evaluated.

    The metric for evaluation will be average precision at IoU=0.5, i.e. AP@0.5, computed from precision-recall curves. The scripts for calculation of this metric will be provided tothe participants in the starting kit.
    For the three tracks, eligible entries received will be judged using the criteria describedabove to determine winners.

  2. Tentative Contest Schedule

    The registered participants will be notified by email of any change in the following tentative schedule:


      • Start of the competition (November 19th, 2019): Beginning of the quantitative competition. We release training (with groundtruth) and validation data (without groundtruth).

      • Release: encrypted test data and validation GT (January 22th, 2020): Release of encrypted test data (without groundtruth) and release of the validation groundtruth.

      • Test phase begins (January 25th, 2020): Release of test data decryption key. Participants start predicting the results on the test data.

      • End of the competition (February 4th, 2020): Deadline for submitting the final predictions over the test (evaluation) data. The organizers start the code verification by running it on the final evaluation data. Deadline for code submission.

      • Submission of fact sheets and material (February 8th, 2020): Deadline for submitting the fact sheets and all the requested material for verification (including code, trained models, configuration files, instructions to verify the code, and so on).

    • Verification of results (February 8th, 2020): Release of the verification results to the participants for review. Participants are invited to follow the paper submission guide for submitting competition papers to the associated FG2020 workshop.

    Optionally:

    • Paper submission deadline (February 15th, 2020): Paper submission deadline for submitting their FG2020 workshop paper.
    • Notification to authors (February 23th, 2020): Notification of acceptance/rejection to authors of the submitted papers.
    • Camera-ready paper submission deadline (February 27th, 2020)
  3. Eligibility

    You are eligible to enter this contest if you meet the following requirements:

      1. You are an individual or a team of people desiring to contribute to the tasks of thechallenge and accepting to follow its rules; you are employed by a research laboratory,startup or other legal entity having a scientific research department or activity ; and

      2. You are NOT a resident of any country constrained by US export regulations in-cluded in the OFAC sanction page http://www.treasury.gov/resource-center/sanctions/Programs/Pages/Programs.aspx. Therefore residents of these countries/ regions are not eligible to participate; and

    You are not an employee of CHALEARN or any of the sponsoring or co-organizingentities; and

        1. You are not involved in any part of the administration and execution of this contest;and

        2. You are not an immediate family (parent, sibling, spouse, or child) or householdmember of an employee of CHALEARN or a person involved in any part of theadministration and execution of this contest.

    This contest is void within the geographic area identified above and wherever else prohibited by law. If you choose to submit an entry, but are not qualified to enter thecontest, this entry is voluntary, and any entry you submit is governed by the remainder ofthese contest rules; CHALEARN reserves the right to evaluate it for scientific purposes. Ifyou are not qualified to submit a contest entry and still choose to submit one, under no circumstances will such entries qualify for sponsored prizes

  4. Entry

    To be eligible for judging, an entry must meet the following content/technical requirements:

    1. Entry contents: The participants are required to submit prediction results and code.To be eligible for prizes, the top ranking participants are required to publicly release their code under a license of their choice, taken among popular OSI-approved licenses ((http://opensource.org/licenses) and make their code accessible on-line for a period of not less than three years following the end of the challenge (only required for top three ranked participants of the competition). To be part of the final ranking the participants will be asked to fill out a survey (fact sheet) briefly describing their method. The top ranking participants and the rest of participants are also invited (not mandatory) to submit a maximum 8-page paper for the proceedings of the associated FG2020 ChaLearn Workshop under evaluation (to be held in May 2020).To be eligible for prizes, top ranked participants score must improve the baseline performance provided by the challenge organizers.

    2. Pre-requisite: There is no pre-requisite to participate, including no requirement to have participated in previous challenges.

    3. Use of data provided: All data provided by CHALEARN are freely available to the participants from the website of the challenge under license terms provided with the data. The data are available only for open research and educational purposes, within the scope of the challenge. ChaLearn and the organizers make no warranties regarding the database, including but not limited to warranties of non-infringement or fitness for a particular purpose. The copyright of the videos remains in propertyof their respective owners. By downloading and making use of the data, you accept full responsibility for using the data. You shall defend and indemnify ChaLearn andthe organizers, including their employees, Trustees, officers and agents, against anyand all claims arising from your use of the data. You agree not to redistribute the data without this notice

      • Test data: The organizers will use test data to perform the final evaluation, hence the participants final entry will be based on test data.

      • Training and validation data: The contest organizers will make available to the participants a training dataset with truth labels, and a validation set with notruth labels. The validation data will be used by the participants for practice purposes to validate their systems. It will be similar in composition to the test set (validation labels may be provided in the final test stage of thechallenge). Post-challenge analyses: The organizers may also perform additional post-challenge analyses using extra data, but the results will not affect the ranking of the challenge performed with the test data.

    4. Submission: The entries of the participants will be submitted on-line via the Codalab web platform. During the development period (quantitative competition), the participants will receive immediate feedback on validation data released for practice purpose. For the final quantitative evaluation, the results will be computed automatically on test data submissions. The performances on test data will not be released until the challenge is over. A jury of experts using test data will perform the qualitative evaluation of explanatory interfaces.

    5. Original work, permissions: In addition, by submitting your entries into this contest you confirm that, to the best of your knowledge:

      • Your entry is your own original work; and

      • Your entry only includes material that you own, or that you have permission from the copyright / trademark owner to use.

  5. Potential use of entry

    Other than what is set forth below, we are not claiming any ownership rights to your entry. However, by submitting your entry, you:

      1. Are granting us an irrevocable, worldwide right and license, in exchange for your opportunity to participate in the contest and potential prize awards, for the duration of the protection of the copyrights to:

        1. Use, review, assess, test and otherwise analyze results submitted or produced by your code and other material submitted by you in connection with this contest and any future research or contests sponsored by; and

        2. Feature your entry and all its content in connection with the promotion of this contest in all media (now known or later developed);

      2. Agree to sign any necessary documentation that may be required for us and our designees to make use of the rights you granted above;

      3. Understand that we cannot control the incoming information you will disclose to our representatives or our co-sponsors representatives in the course of entering, or what our representatives will remember about your entry. You also understand that we will not restrict work assignments of representatives or our co-sponsors representatives who have had access to your entry. By entering this contest, you agree that use of information in our representatives or our co-sponsors representatives unaided memories in the development or deployment of our products or services does not create liability for us under this agreement or copyright or trade secret law;

      4. Understand that you will not receive any compensation or credit for use of your entry,other than what is described in these official rules.

    If you do not want to grant us these rights to your entry, please do not enter this contest.

  6. Submission of entries

    1. Follow the instructions on the Codalab website to submit entries.

    2. The participants will be registered as mutually exclusive teams. Each team may submit only one single final entry. We are not responsible for entries that we do not receive for any reason, or for entries that we receive but are not functioning properly.

    3. The participants must follow the instructions. We will automatically disqualify incomplete or invalid entries.

  7. Judging the entries

    The board of CHALEARN will select a panel of judges to judge the entries; all judges will be forbidden to enter the contest and will be experts in causality, statistics, machine learning, computer vision, or a related field, or experts in challenge organization. A list of the judges will be made available upon request. The judges will review all eligible entries received and select three winners for each of the two tracks. For the quantitative competition winners will be determined based upon the prediction score on test data. For the qualitative track, a jury of experts will evaluate the explanations generated by the interface and the interface itself, the criteria for evaluation are available in the challenge website. The judges will verify that the winners complied with the rules, including that they documented their method by filling out a fact sheet.

    The decisions of these judges are final and binding. The distribution of prizes according to the decisions made by the judges will be made within three (3) months after completion of the last round of the contest. If we do not receive a sufficient number of entries meeting the entry requirements, we may, at our discretion based on the above criteria, not award any or all of the contest prizes below. In the event of a tie between any eligible entries, the tie will be broken by giving preference to the earliest submission, using the time stamp of the submission platform.

  8. Prizes and awards

    1. ChaLearn, University of Barcelona, Computer Vision Center at Autonomous University of Barcelona, and Human Pose Recovery and Behavior Analysis Group are the financial sponsors of this contest. There may be economic incentive prizes and travel grants for the winners (based on availability) to boost contest participation; these prizes will not require participants to enter into an IP agreement with any ofthe sponsors, to disclose algorithms, or to deliver source code to them.

    2. Incentive Prizes for each track

      Award certificates and travel awards (based on availability) will be attributed to the top 3 ranked participants of each track. In addition top ranked participants will beinvited to submit a paper to the associated FG2020 Workshop (pending acceptance).

      (*) The amount of travel awards will be based on need and availability. The travel award may be used for one of the workshops organized in conjunction with the challenge. The award money will be granted in reimbursement of expenses including airfare, ground transportation, hotel, or workshop registration. Reimbursement is conditioned on (i) attending the workshop, (ii) making an oral presentation of the methods used in the challenge, and (iii) presenting original receipts and boardingpasses.

    3. Travel awards: Other travel awards may be distributed to deserving participants based upon need and availability.

    4. If for any reason the advertised prize is unavailable, unless to do so would be prohibited by law, we reserve the right to substitute a prize(s) of equal or greater value, aspermitted. We will only award one prize per team. If you are selected as a potential winner of this contest:

      • If your prize is not in cash, you may not exchange your prize for cash; you may not exchange any prize for other merchandise or services.

      • You may not designate someone else as the winner. If you are unable or unwilling to accept your prize, we will award it to an alternate potential winner.

      • If you accept a prize, you will be solely responsible for all applicable taxes related to accepting the prize.

      • If you are a minor in your place of residence, we may award the prize to your parent/legal guardian on your behalf and your parent/legal guardian will bedesignated as the winner.

  9. Other Sponsored Events

    1. To stimulate participation, the organizers are making available several channels of scientific paper publication. Publishing papers is optional and will not be a conditionto entering the challenge or winning prizes.

    2. The results of the challenge will be presented in the competition program of FG2020. Also, an overview paper will be published in the associated FG2020 ChaLearn Workshop. A selection of the best workshop papers may be invited to submit extended versions of their papers for a special issue in a top tier journal. The results of the challenge will be published in the FG 2017 ChaLearn Workshop on Privacy-aware Computer Vision proceedings. Also, the top ranked participants and/or those developing innovative solutions may be invited to write a joint paper with the organizers to be submitted to a top tier journal.

    The organizers may also sponsor other events to stimulate participation.

  10. Notifications

    If there is any change to data, schedule, instructions of participation, or these rules, the registered participants will be notified at the email they provided with the registration.

    If you are a potential winner, we will notify you by sending a message to the e-mail address listed on your final entry within seven days following the determination of winners. If the notification that we send is returned as undeliverable, or you are otherwise unreachable for any reason, we may award the prize to an alternate winner, unless forbidden by applicable law.

    Winners who have entered the contest as a team will be responsible to share any prize among their members. The prize will be delivered to the registered team leader. If this person becomes unavailable for any reason, the prize will be delivered to be the authorized account holder of the e-mail address used to make the winning entry.

    If you are a potential winner, we may require you to sign a declaration of eligibility, use, indemnity and liability/publicity release and applicable tax forms. If you are a potential winner and are a minor in your place of residence, and we require that your parent or legal guardian will be designated as the winner, and we may require that they sign a declaration of eligibility, use, indemnity and liability/publicity release on your behalf. If you, (or your parent/legal guardian if applicable), do not sign and return these required forms within the time period listed on the winner notification message, we may disqualify you (or the designated parent/legal guardian) and select an alternate selected winner.

  11. On-line notification

    We will post changes in the rules or changes in the data as well as the names of confirmed winners (after contest decisions are made by the judges) online on http://gesture.chalearn.org/ and http://chalearnlap.cvc.uab.es/. This list will remain posted for one year or will bemade available upon request by sending email to mmgesture@chalearn.org

  12. Conditions

    By entering this contest you agree:

    1. To abide by these official rules;

    2. To the extent allowable under applicable law, to release and hold harmless CHALEARN and sponsors, their respective parents, subsidiaries, affiliates, employees and agents from any and all liability or any injury, loss, damage, right, claim or action of any kind arising from or in connection with this contest or any prize won save for residents of the United Kingdom, Chile, Korea, Greece, Brazil, Turkey, Hong Kong,France and Germany with respect to claims resulting from death or personal injury arising from CHALEARNs, Computer Vision Center at Autonomous University of Barcelona, University of Barcelona's negligence, for residents of the United Kingdom with respect to claims resulting from the tort of deceit or any other liabilities that may not be excluded by law, and for residents of Australia in respect of any implied condition or warranty the exclusion of which from these official rules would contravene any statute or cause any part of these official rules to be void;

    3. That CHALEARNs decisions will be final and binding on all matters related to thiscontest; and

    4. That by accepting a prize, CHALEARN and competition sponsors may use your team name, your name, and your place of residence online and in print, or in any other media, in connection with this contest, without payment or compensation to you. The declaration of eligibility, use, indemnity and liability/publicity release provided to the potential winner will make reference to obtaining his/her free consent to use his/her name and place of residence. In any case, the lack of such consent does notprevent the winner from receiving the prize.

    5. This contest will be governed by the laws of the state of California, and you consent to the exclusive jurisdiction and venue of the courts of the state of California forany disputes arising out of this contest. For residents of Austria only: you may withdraw your submission from this contest within seven days of your entry. If you withdraw within seven days of entry, your submission will be returned to you, and we will not make any use of your submission in the future. However, you will notbe eligible to win a prize. If you do not withdraw within seven days of entry, you will be bound by the provisions of these official rules. For residents of the United Kingdom only: the provisions of the contracts (rights of third parties) act 1999 willnot apply to this agreement. For residents of New Zealand only: the provisions of thecontracts (privity) act of 1982 will not apply to this agreement. For Quebec residents: any litigation respecting the conduct or organization of a publicity contest may be submitted to the Régie des Alcools, des Courses et des Jeux for ruling. Any litigation respecting the awarding of a prize may be submitted to the Rgie only for the purpose of helping the parties reach a settlement. For residents of Israel only: this agreement does not entitle third parties to benefits under this agreement as defined in Chapter D of the Contracts Act (General Part) 1973.

    6. The data are available only for research and educational purposes, within the scopeof the challenge. ChaLearn and the organizers make no warranties regarding the database, including but not limited to warranties of non-infringement or fitness for a particular purpose. The copyright of the videos remain the property of their respective owners. By downloading and making use of the data, you accept full responsibility for using the data. You shall defend and indemnify ChaLearn and the organizers, including their employees, Trustees, officers and agents, against any and all claimsarising from your use of the data. You agree not to redistribute the data without this notice.

  13. Unforeseen event

    If an unforeseen or unexpected event (including, but not limited to: someone cheating; a virus, Bug, or catastrophic event corrupting data or the submission platform; someone discovering a flaw in the data or modalities of the challenge) that cannot be reasonably anticipated or controlled, (also referred to as force majeure) affects the fairness and / or integrity of this contest, we reserve the right to cancel, change or suspend this contest. This right is reserved whether the event is due to human or technical error. If a solution cannot be found to restore the integrity of the contest, we reserve the right to select winners basedon the criteria specified above from among all eligible entries received before we had to cancel, change or suspend the contest subject to obtaining the approval from the Régie des Alcools, des Courses et des Jeux with respect to the province of Quebec.

    Computer hacking is unlawful. If you attempt to compromise the integrity or the legitimate operation of this contest by hacking or by cheating or committing fraud in anyway, we may seek damages from you to the fullest extent permitted by law. Further, we may ban you from participating in any of our future contests, so please play fairly.

  14. Sponsor

    ChaLearn is the sponsor of this contest,

    955 Creston Road,
    Berkeley, CA 94708, USA
    events@chalearn.org

    and

    University of Barcelona, Computer Vision Center at Autonomous University of Barcelona, Human Pose Recovery and Behavior Analysis group, are the co-sponsors of this contest. Additional sponsors can be added during the competition period.

  15. Privacy

    During the development phase of the contest and when they submit their final entries, contest participants do not need to disclose their real identity, but must provide a valid email address where we can be deliver notifications to them regarding the contest. To be eligible for prizes, however, contest participants will need to disclose their real identityto contest organizers, informing them by email of their name, professional affiliation, and address. To enter the contest, the participants will need to become users of the Codalab platform. Any profile information stored on this platform can be viewed and edited by the users. After the contest, the participants may cancel their account with the Codalaband cease to be users of that platform. All personal information will then be destroyed. The Codalab privacy policy will apply to contest information submitted by participants on the Codalab. Otherwise, CHALEARNs privacy policy will apply to this contest and to all information that we receive from your entry that we receive directly from you or whichyou have submitted as part of your contest entry on the Codalab. Please read the privacy policy on the contest entry page before accepting the official rules and submitting your entry. Please note that by accepting the official rules you are also accepting the terms of the CHALEARN privacy policy: http://www.chalearn.org/privacy.html.

  16. DISCLAIMER

    ALL INFORMATION, SOFTWARE, DOCUMENTATION, AND DATA ARE PROVIDED ”AS-IS”. THE ORGANIZERS DISCLAIM ANY EXPRESSED OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL CHALEARN AND/OR OTHER ORGANIZERS BE LIABLE FOR ANY SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF SOFTWARE, DOCUMENTS, MATERIALS, PUBLICATIONS, OR INFORMATION MADE AVAILABLE FOR THE CHALLENGE.

 

Learning

Start: Nov. 19, 2019, midnight

Final Evaluation

Start: Jan. 25, 2020, midnight

Competition Ends

April 2, 2020, midnight

You must be logged in to participate in competitions.

Sign In
# Username Score
1 phalanx 0.582473
2 IPHD_baselines 0.541987
3 Mingbo_Hong 0.099885