# The output type of net.forward after blobFromImages **URL:** \ **Category:** C++\ **Tags:** dnn\ **Created:** [September 29, 2024, 9:34am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859 "2024-09-29T09:34:25Z")\ **Posts on this page:** 8\ **Page:** 1
**Author:** ![Sea\_Sea](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/sea_sea/32/11091_2.png) [@Sea\_Sea](https://forum.opencv.org/u/Sea_Sea)\ **Post date:** [September 29, 2024, 9:34am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/1 "2024-09-29T09:34:25Z")
Assume zidane.jpg size is (480, 640, 3) The case1 network output is 1 x 84 x 6300 where the 1 position is the batch (Batch, Classes, Boxes) But the case2 network output is 1 x 84 x 6300 for outs[0] How to define the i of outs[i] (std::vectorcv::Mat)? Is `i` a batch? but the outs[i] shape is (Batch, Classes, Boxes) two batch? Because the input of `cv::dnn::blobFromImages` can be the `std::vector` where the j of inpMats[j] can be the batch. It just like inpMats[0] : (480, 640, 3) inpMats[1] : (480, 640, 3) case1 . ```auto std::vector inpMats; // inpMats.push_back(cv::imread("source/data/zidane.jpg")); inpMats.push_back(cv::imread("source/data/zidane.jpg")); cv::Mat blob1 = cv::dnn::blobFromImages(inpMats, 1.0 / 255.0, cv::Size(640, 480), cv::Scalar(), true, false); // blob1 Dimensions: 1 x 3 x 480 x 640 net.setInput(blob1); cv::Mat single_out = net.forward(); // single_out Dimensions: 1 x 84 x 6300 ``` case2 ```auto std::vector inpMats; // inpMats.push_back(cv::imread("source/data/zidane.jpg")); inpMats.push_back(cv::imread("source/data/zidane.jpg")); cv::Mat blob1 = cv::dnn::blobFromImages(inpMats, 1.0 / 255.0, cv::Size(640, 480), cv::Scalar(), true, false); // blob1 Dimensions: 1 x 3 x 480 x 640 net.setInput(blob1); std::vector outs; net.forward(outs, net.getUnconnectedOutLayersNames()); // outs Dimensions: 1 x 84 x 6300 for outs[0] ``` ---
**Author:** ![berak](https://avatars.discourse-cdn.com/v4/letter/b/85f322/32.png) [@berak](https://forum.opencv.org/u/berak)\ **Post date:** [September 29, 2024, 9:37am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/2 "2024-09-29T09:37:21Z")
what kind of network is it ? what are you trying to achieve ? (the _purpose_ of it ?) what does `net.getUnconnectedOutLayersNames()` return ? please, try to make [MRE](https://stackoverflow.com/help/minimal-reproducible-example) from your code snippets, so ppl here _can_ actually try to reproduce it, ty. ---
**Author:** ![Sea\_Sea](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/sea_sea/32/11091_2.png) [@Sea\_Sea](https://forum.opencv.org/u/Sea_Sea)\ **Post date:** [September 29, 2024, 9:47am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/3 "2024-09-29T09:47:02Z")
I just confuse about the case 2 network output define ( std::vectorcv::Mat outs.) If the outs is defined (Batch, Classes, Boxes), what is the i of outs[i] I already know the std::vectorcv::Mat inpMat define is (480, 640, 3) for inpMat[j] and the j is the batch I reference these code and I want to design the input (many img) and output (batch output), > [opencv/modules/dnn/test/test\_onnx\_importer.cpp at 450e741f8d53ff12b4e194c7762adaefb952555a · opencv/opencv · GitHub](https://github.com/opencv/opencv/blob/450e741f8d53ff12b4e194c7762adaefb952555a/modules/dnn/test/test_onnx_importer.cpp#L2249) # Reproduce Just follow the code that I paste and I use blobFromImages ```auto std::vector inpMats; inpMats.push_back(cv::imread("source/data/zidane.jpg")); cv::Mat blob1 = cv::dnn::blobFromImages(inpMats, 1.0 / 255.0, cv::Size(640, 480), cv::Scalar(), true, false); net.setInput(blob1); //case1 cv::Mat single_out = net.forward(); // single_out Dimensions: 1 x 84 x 6300 // case2 std::vector outs;. net.forward(outs); //net.forward(outs, net.getUnconnectedOutLayersNames()); // outs Dimensions: 1 x 84 x 6300 for outs[0] ``` ---
**Author:** ![berak](https://avatars.discourse-cdn.com/v4/letter/b/85f322/32.png) [@berak](https://forum.opencv.org/u/berak)\ **Post date:** [September 29, 2024, 10:42am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/4 "2024-09-29T10:42:48Z")
sorry, your reproducer code is incomplete (no network loaded) however, IF it is really yolov8, we know, that it has a single output layer only, thus you only need to check `outputs[0]` (and it’s the same as a simple `net.forward()`) (other yolo nn’s have 2 or 3 outputs (pyramid scale levels), which you have to collect, before applying NMS) ---
**Author:** ![Sea\_Sea](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/sea_sea/32/11091_2.png) [@Sea\_Sea](https://forum.opencv.org/u/Sea_Sea)\ **Post date:** [September 29, 2024, 10:59am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/5 "2024-09-29T10:59:17Z")
Thanks for your reply I think the yolov8 code is simply, so I just point out the place that I want to focus. According to your reply, you mean that if I choose to use the net output from case 2, I only need to focus on outs[0], The net output will always only output outs[0], and the batch is already included in outs[0]. ? And the net output is impossible for 1 x 84 x 6300 for outs[1] ```auto // case2 std::vector outs;. net.forward(outs); //net.forward(outs, net.getUnconnectedOutLayersNames()); // outs Dimensions: 1 x 84 x 6300 for outs[0] // or Dimensions: 10 x 84 x 6300 for outs[0] // it is wrong : 1 x 84 x 6300 for outs[1] impossible ``` ---
**Author:** ![berak](https://avatars.discourse-cdn.com/v4/letter/b/85f322/32.png) [@berak](https://forum.opencv.org/u/berak)\ **Post date:** [September 29, 2024, 11:34am UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/6 "2024-09-29T11:34:54Z")
ok, so part 2, - batches : a batch of 7 input images will result in an input blob like ``` [7, 3, 480, 640] // NCHW ``` and (again, for v8, a _single_) nn output like: ``` [7, 84, 6300] // batch, box proposal, count ``` so you still have to parse / process box proposals _per image_. maybe the short answer is: the number/index of output layers is per scale level, not the image/batch count (which is the 1st index of the resp. output layer tensor) ---
**Author:** ![Sea\_Sea](https://sea2.discourse-cdn.com/flex020/user_avatar/forum.opencv.org/sea_sea/32/11091_2.png) [@Sea\_Sea](https://forum.opencv.org/u/Sea_Sea)\ **Post date:** [September 29, 2024, 12:56pm UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/7 "2024-09-29T12:56:06Z")
Yes , you are right. ### conclusion yolov8 output is one output layer tensor, so there is only outs[0] for case 2 design If some network output is three output layer tensor, there are three (outs[0] outs[1] outs[2]) for case 2 design. ### new question But if some network output is three output layer tensor, it can not use the case 1 design, right? ```auto case 1 network output type is cv::Mat case 2 network output type is std::vector ``` ---
**Author:** ![berak](https://avatars.discourse-cdn.com/v4/letter/b/85f322/32.png) [@berak](https://forum.opencv.org/u/berak)\ **Post date:** [September 29, 2024, 1:01pm UTC](https://forum.opencv.org/t/the-output-type-of-net-forward-after-blobfromimages/18859/8 "2024-09-29T13:01:55Z")
> [@Sea\_Sea](#): > > But if some network output is three output layer tensor, it can not use the case 1 design, right > > ```auto > > ``` indeed ! there are a lot of nn’s with multiple outputs, e.g. pose nn’s (with keypoint / connection outputs) or even yolo variants with additional segmentation maps