標簽分配（label assignment）

什么是標簽分配

標簽分配（Label Assignment）標簽分配策略是對訓(xùn)練過程中各個Anchor劃分正負屬性，并分配各自學(xué)習(xí)目標的策略方法，在整體上通過標簽是否是非負即正可以分為硬標簽分配和軟標簽分配。其中，硬標簽分配可以分成靜態(tài)分配策略和動態(tài)分配策略兩類。
動態(tài)

• 靜態(tài)分配策略
  靜態(tài)標簽分配方法主要基于距離、IOU等先驗知識設(shè)置固定閾值去區(qū)分正負樣本，如FCOS、兩階段標檢測算法、RFLA等。
• 動態(tài)分配策略
  動態(tài)標簽分配方法則根據(jù)不同策略動態(tài)設(shè)置閾值選擇正負樣本，如ATSS、PAA、OTA、DSL、SimOTA等。

OTA損失函數(shù)介紹

OTA（Optimal Transport Assignment for Object Detection）出處為2021年 CVPR會議上的一篇論文。該論文的貢獻主要有以下兩點：

1. 提出了一種基于優(yōu)化策略的標簽分配方式，將 gt 看做 label 供應(yīng)商，anchor 看做 label 需求方。對于正樣本，將分類和回歸的 loss 加權(quán)和作為傳輸花費，對于負樣本，傳輸花費就為分類 loss，通過最小化該花費，讓網(wǎng)絡(luò)自己學(xué)習(xí)最優(yōu)的標簽分配方式。
2. 免去了手工選定參數(shù)的方式來實現(xiàn)標簽分配，讓網(wǎng)絡(luò)自己選擇每個 gt 對應(yīng)的 anchor 數(shù)量，而非提前設(shè)定，也能夠較好的解決模棱兩可的 anchor 分配問題，提高網(wǎng)絡(luò)對這部分 anchor 的處理效果

論文：Optimal Transport Assignment for Object Detection
代碼：OTA代碼

背景

Label assignment 在object detection領(lǐng)域中發(fā)揮著非常重要的作用，它能夠分配每個 anchor 的正負。但是傳統(tǒng)的方法具有一些缺點，具體體現(xiàn)為：不同大小、形狀、遮擋程度的目標，其 positive/negative 的判定條件應(yīng)該是不同的。針對這個缺點，就有一些方法使用了動態(tài)分配的策略來實現(xiàn)label assignment。

方法

這個方法可以看這個博客，講的比較詳細：博客文章來源地址http://www.zghlxwxcb.cn/news/detail-649180.html

如何在yolo v5目標檢測算法中改為OTA損失

步驟一、修改loss.py文件

import torch.nn.functional as F
from utils.metrics import box_iou
from utils.torch_utils import de_parallel
from utils.general import xywh2xyxy

class ComputeLossOTA:
    # Compute losses
    def __init__(self, model, autobalance=False):
        super(ComputeLossOTA, self).__init__()
        device = next(model.parameters()).device  # get model device
        h = model.hyp  # hyperparameters

        # Define criteria
        BCEcls = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([h['cls_pw']], device=device))
        BCEobj = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([h['obj_pw']], device=device))

        # Class label smoothing https://arxiv.org/pdf/1902.04103.pdf eqn 3
        self.cp, self.cn = smooth_BCE(eps=h.get('label_smoothing', 0.0))  # positive, negative BCE targets

        # Focal loss
        g = h['fl_gamma']  # focal loss gamma
        if g > 0:
            BCEcls, BCEobj = FocalLoss(BCEcls, g), FocalLoss(BCEobj, g)

        det = de_parallel(model).model[-1]  # Detect() module
        self.balance = {3: [4.0, 1.0, 0.4]}.get(det.nl, [4.0, 1.0, 0.25, 0.06, .02])  # P3-P7
        self.ssi = list(det.stride).index(16) if autobalance else 0  # stride 16 index
        self.BCEcls, self.BCEobj, self.gr, self.hyp, self.autobalance = BCEcls, BCEobj, 1.0, h, autobalance
        for k in 'na', 'nc', 'nl', 'anchors', 'stride':
            setattr(self, k, getattr(det, k))

    def __call__(self, p, targets, imgs):  # predictions, targets, model   
        device = targets.device
        lcls, lbox, lobj = torch.zeros(1, device=device), torch.zeros(1, device=device), torch.zeros(1, device=device)
        bs, as_, gjs, gis, targets, anchors = self.build_targets(p, targets, imgs)
        pre_gen_gains = [torch.tensor(pp.shape, device=device)[[3, 2, 3, 2]] for pp in p] 
    

        # Losses
        for i, pi in enumerate(p):  # layer index, layer predictions
            b, a, gj, gi = bs[i], as_[i], gjs[i], gis[i]  # image, anchor, gridy, gridx
            tobj = torch.zeros_like(pi[..., 0], device=device)  # target obj

            n = b.shape[0]  # number of targets
            if n:
                ps = pi[b, a, gj, gi]  # prediction subset corresponding to targets

                # Regression
                grid = torch.stack([gi, gj], dim=1)
                pxy = ps[:, :2].sigmoid() * 2. - 0.5
                #pxy = ps[:, :2].sigmoid() * 3. - 1.
                pwh = (ps[:, 2:4].sigmoid() * 2) ** 2 * anchors[i]
                pbox = torch.cat((pxy, pwh), 1)  # predicted box
                selected_tbox = targets[i][:, 2:6] * pre_gen_gains[i]
                selected_tbox[:, :2] -= grid
                iou = bbox_iou(pbox, selected_tbox, CIoU=True)  # iou(prediction, target)
                if type(iou) is tuple:
                    lbox += (iou[1].detach() * (1 - iou[0])).mean()
                    iou = iou[0]
                else:
                    lbox += (1.0 - iou).mean()  # iou loss

                # Objectness
                tobj[b, a, gj, gi] = (1.0 - self.gr) + self.gr * iou.detach().clamp(0).type(tobj.dtype)  # iou ratio

                # Classification
                selected_tcls = targets[i][:, 1].long()
                if self.nc > 1:  # cls loss (only if multiple classes)
                    t = torch.full_like(ps[:, 5:], self.cn, device=device)  # targets
                    t[range(n), selected_tcls] = self.cp
                    lcls += self.BCEcls(ps[:, 5:], t)  # BCE

                # Append targets to text file
                # with open('targets.txt', 'a') as file:
                #     [file.write('%11.5g ' * 4 % tuple(x) + '\n') for x in torch.cat((txy[i], twh[i]), 1)]

            obji = self.BCEobj(pi[..., 4], tobj)
            lobj += obji * self.balance[i]  # obj loss
            if self.autobalance:
                self.balance[i] = self.balance[i] * 0.9999 + 0.0001 / obji.detach().item()

        if self.autobalance:
            self.balance = [x / self.balance[self.ssi] for x in self.balance]
        lbox *= self.hyp['box']
        lobj *= self.hyp['obj']
        lcls *= self.hyp['cls']
        bs = tobj.shape[0]  # batch size

        loss = lbox + lobj + lcls
        return loss * bs, torch.cat((lbox, lobj, lcls)).detach()

    def build_targets(self, p, targets, imgs):
        indices, anch = self.find_3_positive(p, targets)
        device = torch.device(targets.device)
        matching_bs = [[] for pp in p]
        matching_as = [[] for pp in p]
        matching_gjs = [[] for pp in p]
        matching_gis = [[] for pp in p]
        matching_targets = [[] for pp in p]
        matching_anchs = [[] for pp in p]
        
        nl = len(p)    
    
        for batch_idx in range(p[0].shape[0]):
        
            b_idx = targets[:, 0]==batch_idx
            this_target = targets[b_idx]
            if this_target.shape[0] == 0:
                continue
                
            txywh = this_target[:, 2:6] * imgs[batch_idx].shape[1]
            txyxy = xywh2xyxy(txywh)

            pxyxys = []
            p_cls = []
            p_obj = []
            from_which_layer = []
            all_b = []
            all_a = []
            all_gj = []
            all_gi = []
            all_anch = []
            
            for i, pi in enumerate(p):
                
                b, a, gj, gi = indices[i]
                idx = (b == batch_idx)
                b, a, gj, gi = b[idx], a[idx], gj[idx], gi[idx]                
                all_b.append(b)
                all_a.append(a)
                all_gj.append(gj)
                all_gi.append(gi)
                all_anch.append(anch[i][idx])
                from_which_layer.append((torch.ones(size=(len(b),)) * i).to(device))
                
                fg_pred = pi[b, a, gj, gi]                
                p_obj.append(fg_pred[:, 4:5])
                p_cls.append(fg_pred[:, 5:])
                
                grid = torch.stack([gi, gj], dim=1)
                pxy = (fg_pred[:, :2].sigmoid() * 2. - 0.5 + grid) * self.stride[i] #/ 8.
                #pxy = (fg_pred[:, :2].sigmoid() * 3. - 1. + grid) * self.stride[i]
                pwh = (fg_pred[:, 2:4].sigmoid() * 2) ** 2 * anch[i][idx] * self.stride[i] #/ 8.
                pxywh = torch.cat([pxy, pwh], dim=-1)
                pxyxy = xywh2xyxy(pxywh)
                pxyxys.append(pxyxy)
            
            pxyxys = torch.cat(pxyxys, dim=0)
            if pxyxys.shape[0] == 0:
                continue
            p_obj = torch.cat(p_obj, dim=0)
            p_cls = torch.cat(p_cls, dim=0)
            from_which_layer = torch.cat(from_which_layer, dim=0)
            all_b = torch.cat(all_b, dim=0)
            all_a = torch.cat(all_a, dim=0)
            all_gj = torch.cat(all_gj, dim=0)
            all_gi = torch.cat(all_gi, dim=0)
            all_anch = torch.cat(all_anch, dim=0)
        
            pair_wise_iou = box_iou(txyxy, pxyxys)

            pair_wise_iou_loss = -torch.log(pair_wise_iou + 1e-8)

            top_k, _ = torch.topk(pair_wise_iou, min(10, pair_wise_iou.shape[1]), dim=1)
            dynamic_ks = torch.clamp(top_k.sum(1).int(), min=1)

            gt_cls_per_image = (
                F.one_hot(this_target[:, 1].to(torch.int64), self.nc)
                .float()
                .unsqueeze(1)
                .repeat(1, pxyxys.shape[0], 1)
            )

            num_gt = this_target.shape[0]
            cls_preds_ = (
                p_cls.float().unsqueeze(0).repeat(num_gt, 1, 1).sigmoid_()
                * p_obj.unsqueeze(0).repeat(num_gt, 1, 1).sigmoid_()
            )

            y = cls_preds_.sqrt_()
            pair_wise_cls_loss = F.binary_cross_entropy_with_logits(
               torch.log(y/(1-y)) , gt_cls_per_image, reduction="none"
            ).sum(-1)
            del cls_preds_
        
            cost = (
                pair_wise_cls_loss
                + 3.0 * pair_wise_iou_loss
            )

            matching_matrix = torch.zeros_like(cost, device=device)

            for gt_idx in range(num_gt):
                _, pos_idx = torch.topk(
                    cost[gt_idx], k=dynamic_ks[gt_idx].item(), largest=False
                )
                matching_matrix[gt_idx][pos_idx] = 1.0

            del top_k, dynamic_ks
            anchor_matching_gt = matching_matrix.sum(0)
            if (anchor_matching_gt > 1).sum() > 0:
                _, cost_argmin = torch.min(cost[:, anchor_matching_gt > 1], dim=0)
                matching_matrix[:, anchor_matching_gt > 1] *= 0.0
                matching_matrix[cost_argmin, anchor_matching_gt > 1] = 1.0
            fg_mask_inboxes = (matching_matrix.sum(0) > 0.0).to(device)
            matched_gt_inds = matching_matrix[:, fg_mask_inboxes].argmax(0)
        
            from_which_layer = from_which_layer[fg_mask_inboxes]
            all_b = all_b[fg_mask_inboxes]
            all_a = all_a[fg_mask_inboxes]
            all_gj = all_gj[fg_mask_inboxes]
            all_gi = all_gi[fg_mask_inboxes]
            all_anch = all_anch[fg_mask_inboxes]
        
            this_target = this_target[matched_gt_inds]
        
            for i in range(nl):
                layer_idx = from_which_layer == i
                matching_bs[i].append(all_b[layer_idx])
                matching_as[i].append(all_a[layer_idx])
                matching_gjs[i].append(all_gj[layer_idx])
                matching_gis[i].append(all_gi[layer_idx])
                matching_targets[i].append(this_target[layer_idx])
                matching_anchs[i].append(all_anch[layer_idx])

        for i in range(nl):
            if matching_targets[i] != []:
                matching_bs[i] = torch.cat(matching_bs[i], dim=0)
                matching_as[i] = torch.cat(matching_as[i], dim=0)
                matching_gjs[i] = torch.cat(matching_gjs[i], dim=0)
                matching_gis[i] = torch.cat(matching_gis[i], dim=0)
                matching_targets[i] = torch.cat(matching_targets[i], dim=0)
                matching_anchs[i] = torch.cat(matching_anchs[i], dim=0)
            else:
                matching_bs[i] = torch.tensor([], device='cuda:0', dtype=torch.int64)
                matching_as[i] = torch.tensor([], device='cuda:0', dtype=torch.int64)
                matching_gjs[i] = torch.tensor([], device='cuda:0', dtype=torch.int64)
                matching_gis[i] = torch.tensor([], device='cuda:0', dtype=torch.int64)
                matching_targets[i] = torch.tensor([], device='cuda:0', dtype=torch.int64)
                matching_anchs[i] = torch.tensor([], device='cuda:0', dtype=torch.int64)

        return matching_bs, matching_as, matching_gjs, matching_gis, matching_targets, matching_anchs           

    def find_3_positive(self, p, targets):
        # Build targets for compute_loss(), input targets(image,class,x,y,w,h)
        na, nt = self.na, targets.shape[0]  # number of anchors, targets
        indices, anch = [], []
        gain = torch.ones(7, device=targets.device).long()  # normalized to gridspace gain
        ai = torch.arange(na, device=targets.device).float().view(na, 1).repeat(1, nt)  # same as .repeat_interleave(nt)
        targets = torch.cat((targets.repeat(na, 1, 1), ai[:, :, None]), 2)  # append anchor indices

        g = 0.5  # bias
        off = torch.tensor([[0, 0],
                            [1, 0], [0, 1], [-1, 0], [0, -1],  # j,k,l,m
                            # [1, 1], [1, -1], [-1, 1], [-1, -1],  # jk,jm,lk,lm
                            ], device=targets.device).float() * g  # offsets

        for i in range(self.nl):
            anchors = self.anchors[i]
            gain[2:6] = torch.tensor(p[i].shape)[[3, 2, 3, 2]]  # xyxy gain

            # Match targets to anchors
            t = targets * gain
            if nt:
                # Matches
                r = t[:, :, 4:6] / anchors[:, None]  # wh ratio
                j = torch.max(r, 1. / r).max(2)[0] < self.hyp['anchor_t']  # compare
                # j = wh_iou(anchors, t[:, 4:6]) > model.hyp['iou_t']  # iou(3,n)=wh_iou(anchors(3,2), gwh(n,2))
                t = t[j]  # filter

                # Offsets
                gxy = t[:, 2:4]  # grid xy
                gxi = gain[[2, 3]] - gxy  # inverse
                j, k = ((gxy % 1. < g) & (gxy > 1.)).T
                l, m = ((gxi % 1. < g) & (gxi > 1.)).T
                j = torch.stack((torch.ones_like(j), j, k, l, m))
                t = t.repeat((5, 1, 1))[j]
                offsets = (torch.zeros_like(gxy)[None] + off[:, None])[j]
            else:
                t = targets[0]
                offsets = 0

            # Define
            b, c = t[:, :2].long().T  # image, class
            gxy = t[:, 2:4]  # grid xy
            gwh = t[:, 4:6]  # grid wh
            gij = (gxy - offsets).long()
            gi, gj = gij.T  # grid xy indices

            # Append
            a = t[:, 6].long()  # anchor indices
            indices.append((b, a, gj.clamp_(0, gain[3] - 1), gi.clamp_(0, gain[2] - 1)))  # image, anchor, grid indices
            anch.append(anchors[a])  # anchors

        return indices, anch

步驟二、在train.py和val.py中修改conpute_loss

• from utils.loss import ComputeLoss 改為ComputeLossOTA。
• 將 compute_loss = ComputeLoss(model) 中的ComputeLoss方法改為 ComputeLossOTA。
• loss, loss_items = compute_loss(pred, targets.to(device)) 括號內(nèi)添加imgs。
• 在val.py中做出和如上所示的修改。
• 運行train.py即可進行訓(xùn)練過程。

到了這里，關(guān)于目標檢測改進系列1：yolo v5網(wǎng)絡(luò)中OTA損失函數(shù)替換的文章就介紹完了。如果您還想了解更多內(nèi)容，請在右上角搜索TOY模板網(wǎng)以前的文章或繼續(xù)瀏覽下面的相關(guān)文章，希望大家以后多多支持TOY模板網(wǎng)！